Globalization: A Woman's Best Friend? Exporters and The Gender Wage Gap

ERIA-DP-2015-25
ERIA Discussion Paper Series
Globalization: A Womans Best Friend? Exporters and

the Gender Wage Gap
Esther Ann BLER
University of Oslo and ESOP
Beata JAVORCIK
University of Oxford, ESOP and CEPR
Karen Helene ULLTVEI-MOE

University of Oslo, ESOP and CEPR
March 2015
Abstract: While the impact of globalization on income inequality has received a lot
of attention, little is known about its effect on the gender wage gap (GWG). This
study argues that there is a systematic differece in GWG between exporting firms
and non-exporters. By the virtue of being exposed to higher competition, exporters
require greater commitment and flexibility from their employees. If commitment is
not easily observable and women are precieved as less committed workers than
men, exporters will statistically discriminate against female employees and will
exhibit a higher GWG than non-exporters. We test this hyphesis using matched
employer-employee data from the Norwegian manufacturing sector from 1996 to
2010. Our identification strategy relies on an exogeneous shock, namely, the
legislative changes that increased the lenght of the parental leave that is available
only to fathers. We argue that these changes have narrowed the perceived
commitment gap between the genders and show that the initially higher GWG
observved in exporting firms relative to non-exporters has gone down after the
changes took place.
Keywords: Exporters, Globalization, Gender Wage Gap

JEL Classification: F10; F14; F16; J16
Globalization: A Womans Best Friend?

Exporters and the Gender Wage Gap
Esther Ann Bler
Beata Javorcik
Karen Helene Ulltveit-Moe
February 2015
Abstract
While the impact of globalization on income inequality has received a lot of attention,
little is known about its effect on the gender wage gap (GWG). This study argues that
there is a systematic difference in the GWG between exporting firms and non-exporters. By
the virtue of being exposed to higher competition, exporters require greater commitment
and flexibility from their employees. If commitment is not easily observable and women
are perceived as less committed workers than men, exporters will statistically discriminate
against female employees and will exhibit a higher GWG than non-exporters. We test this
hypothesis using matched employer-employee data from the Norwegian manufacturing sector
from 1996 to 2010. Our identification strategy relies on an exogenous shock, namely, the
legislative changes that increased the length of the parental leave that is available only to
fathers. We argue that these changes have narrowed the perceived commitment gap between
the genders and show that the initially higher GWG observed in exporting firms relative to
non-exporters has gone down after the changes took place.
JEL: F10, F14, F16, J16.
Keywords: Exporters, Globalization, Gender Wage Gap.
We thank seminar participants at the University of Bergen, University of Oxford, University of Cambridge,
Central European University, Center for Global Development in London, and conference participants at the
European Research Workshop in International Trade (ERWIT) 2014, 10th Joint ECB/CEPR Labour Market
Workshop, ETSG 2013, Annual Meeting of the Norwegian Association of Economists 2014, GEP Post-graduate
Conference 2014, Nordic Workshop on Matched Employer-Employee Data 2014, Nordic International Trade
Seminar (NOITS) 2014 conference and SNF Sinergia/CEPR Conference on Economic Inequality, Labor Markets
and International Trade (ILMIT) for helpful comments. We thank Stefania Albanesi, Andy Bernard, Juan J.
Dolado, Andrea Ichinio, Kjell Erik Lommerud, Monika Merz, Peter Neary, Mari Rege, Olga Solleder, Kjetil
Storesletten and Tony Venables for valuable discussion. This paper is part of the research activities at the Centre
for the Study of Equality, Social Organization, and Performance (ESOP) at the Department of Economics at
the University of Oslo. ESOP is supported by the Research Council of Norway. Data received from Statistics
Norway has been essential for the project and for the paper.
University of Oslo and ESOP.
University of Oxford, ESOP and CEPR.
University of Oslo, ESOP and CEPR.
Globalization: A Womans best friend? Exporters and the Gender Wage Gap
Introduction
The link between globalization and income inequality has received a lot of attention in economic
research as well as in the popular press. However, the debate has almost completely ignored one
important dimension of inequality, namely the gender wage gap. It is a well-documented fact
that women earn less than men, even in relatively equal societies such as the Northern European
countries. This appears to be true even after controlling for observable worker characteristics,
hours worked and occupation, and especially so in the private sector.1 In this paper, we set out
to explore one aspect of this issue by focusing on the link between exporting and the gender
wage gap.
We argue that by the virtue of being exposed to higher competition, exporters require greater
commitment to work and greater flexibility of their employees. For instance, working for an exporting firm may require taking late night phone calls to communicate with customers in different
time zones and may involve international travel arranged at a short notice. The employees may
be expected to be available around the clock seven days a week in case of unexpected problems arising on the production line or shipments being delayed. If commitment is not easily
observable and women are perceived to be less committed workers than men, exporters will
statistically discriminate against female employees and will exhibit a higher gender wage gap
than non-exporters. We expect this to be particularly true among highly skilled employees.2
To investigate the link between the gender wage gap and exporting, we exploit a matched
employer-employee data set covering close to all firms in the Norwegian manufacturing sector and
their full-time employees between 1996 and 2010.3 The data set includes detailed information
on employees, such as number of years of education, labor market experience, gender and other
demographics.
We estimate a Mincerian wage regression controlling for a host of worker and firm characteristics. We find evidence of a substantial overall gender wage gap of around 24%. The gender
wage gap appears to be smaller in exporting firms where it is equal to 19%. Thus working for
an exporting firm is associated with closing almost a quarter of the observed gender wage gap.
These regressions, however, do not take into account unobservable worker characteristics that
are likely to be correlated with their choice of working for an exporter. Interestingly, controlling
for unobservable worker characteristics (or unobservable worker-firm heterogeneity) reverses our
1
See Blau and Kahn (2000), Barth et al. (2013) and Goldin (2014) for overviews.
Employer surveys from Iceland reveal that women are perceived as more family oriented than men and
less committed to their work and less reliable than men (see Gislason (2007)). A third of managers surveyed in the UK claim that women are not as good at their jobs when they come back from maternity leave (see http://www.slatergordon.co.uk/media-centre/news/2014/08/slater-gordon-highlights-maternitydiscrimination). In an experiment, people were asked to evaluate a professional profile which varied in terms of
gender and whether or not the candidate had a child. When identified as a mother, as opposed to a father or
a person whose parental status wasnt mentioned, the profile was judged to be significantly less competent and
was least likely to be hired or promoted by the participants. The mere mention of a child led people to see the
mother as less competent (see Cuddy and Glick (2004)).
3
The data set covers all joint stock firms in the manufacturing sector, and thereby about 90% of manufacturing
output in Norway in 2004.
2
results. We find that exporting firms exhibit a higher gender wage gap than non-exporters with
the difference reaching about 3 percentage points. As expected, this effect is present only among
college graduates. In other words, college educated women earn higher wages at exporting firms
than at non-exporters, but they are underpaid given their unobservable characteristics.
Next we exploit an exogenous shock that is believed to have changed the perceived commitment gap between the genders. The shock consisted of a series of legislative changes that
gradually increased the number of weeks of parental leave that are available only to the childs
father, the so called paternal quota. Unless special circumstances exist, the paternal quota
weeks are lost if not taken by the father. A quota of 4 weeks was first introduced in 1993. It
was extended to 5 and 6 weeks in 2005 and 2006, respectively. In 2009, the number of weeks
was further extended to 10. This nudge on the part of the Norwegian government has resulted
in a huge increase in the fraction of fathers taking a substantial period of paternity leave.
In our empirical analysis, we proxy for the change in social attitudes, and thus the perceived
commitment gap between men and women, using the fraction of fathers taking at least 8 weeks
of leave in a given year. We choose 8 weeks as our cutoff because a two-month long absence
from work is substantial and disruptive enough to affect the employers perceptions. The share of
fathers taking at least 8 weeks of leave went up from 8.4% in 1996 (the first year of our sample),
to 10.8% in 2003 and 40.6% in 2010 (the last year of our sample).4 Our variable of interest is
then a triple interaction between the female dummy, the exporter dummy and the fraction of
fathers taking at least 8 weeks of leave. It captures how the gender wage gap differential between
exporters and non-exporters evolved with the changes in social attitudes.
The estimation results confirm our priors. We find that the gender wage gap is higher
in exporting firms than in non-exporters, and the difference between the two narrows down
with the changes in social attitudes. The estimated coefficients are statistically significant and
economically meaningful. This is true in the full sample as well as in the subsamples of workers
with and without college education. As one would expect, the results are stronger when we focus
our attention on workers in their reproductive years (i.e., those under 45 years of age). They
are also stronger for college graduates.
The magnitude of the estimated effects is economically meaningful. The change in the
fraction of fathers taking at least 8 weeks of paternity leave between 1996 and 2008, corresponds
to the gender wage gap in exporters dropping from being on average 4.5 percent higher than
in other firms to being only 3.3 percent higher.5 For the subsample of workers without college
education, the corresponding gap of 1.3 percent is driven all the way down to zero.
We also exploit an alternative shock in our identification strategy. In 2003, the Norwegian
government initiated an ambitious expansion of kindergarten coverage. The goal of this policy
was to offer all children a high quality, low price place in a public or private kindergarten. It
4
Note that fathers have always been entitled to take more leave than the number of weeks specified in the
paternal quota. The legislation has stipulated a paternal and a maternal quota (the latter equal to 9 weeks since
1993) and has let the parents choose how to split the remainder of the leave.
5
The cited results are based on estimates for college educated workers under 45 years of age.
was also seen as a measure to increase gender equality through the provision of reliable and
formal child care. The initiative resulted in a massive increase in the share of children attending
kindergarten. 9 of 10 children between the ages of 1 and 5 attended kindergarten in 2009, as
compared to 54% in 1996 and 63% in 2001. The expansion of kindergarten coverage proceeded
at different speeds in different municipalities, providing us with geographical variation as well
as time variation in coverage rates. In our estimation, we examine whether the gender wage
gap differential between exporters and non-exporters was affected by the increase in availability
of kindergarten places in a given municipality. The logic behind this approach is that greater
availability of reliable child care is associated with fewer disruptions to work hours and thus
allows women to exhibit greater commitment to their jobs. Our results are consistent with those
found for the paternity leave case. The gender wage gap differential between exporters and
non-exporters decreased as a larger share of children was enrolled in kindergarten. Again the
observed effect was stronger for workers under 45 years of age.
While the first part of the analysis focuses on the perceived commitment gap between the
genders, the second part of the paper exploits heterogeneity among exporters. As firms exporting
to a large number of locations tend to be larger and more productive than other exporters, it
is also likely that they have higher expectations vis a vis their employees, which we expect to
result in them having an even larger gender wage gap.6 This appears to be the case in our data.
We find that the gender wage gap in an exporter increases as the number of export markets
increase.
In the final part of the paper, we turn to a possible alternative interpretation of our baseline
findings of a higher gender wage gap among exporting firms. Beckers theory of taste-based
discrimination predicts that more profitable firms are better positioned to engage in costly
discrimination (Becker (1957)). The literature on trade and heterogeneous firms emphasizes
that exporters are more profitable than other firms, and thus in this framework, exporters may
find it easier to discriminate.7 To address this issue we augment our model by allowing the
gender wage gap to differ with firm profitability (as well as other proxies for profitability, such
as the multinational status and size). Doing so does not affect our finding that the gender wage
gap is higher in exporting firms. Our results indicating that the difference in the gender wage gap
between exporters and non-exporters declined with the changes in gender-related social norms
is also robust to controlling for profitability.
In summary, our findings are consistent with exporters being able to attract more able,
more ambitious or more committed females, whom they subject to a larger degree of statistical
discrimination than do other firms. Our results are consistent with this phenomenon being
driven by exporters attaching greater importance to worker commitment due to operating in
more competitive and fast changing environments than other firms.
This paper speaks to three strands of the existing literature. The first strand is the large
6
See Bernard et al. (2011) and Eaton et al. (2011) for empirical and theoretical work on the relationship
between firm productivity and the number of export destinations.
7
For theoretical prediction see Melitz (2003), for empirical evidence Bernard and Jensen (1999).
literature on exporters and the exporter wage premium. Bernard et al. (1995) show that exporters pay higher wages than non-exporters in the US and that this wage premium goes to
both production and non-production workers. Several papers confirm that this pattern holds in
a host of other countries (see e.g. Schank et al. (2007)). More recently, the availability of matched
employer-employee data has allowed researchers to examine whether the exporter wage premium
is robust to controlling for various worker and firm characteristics. Schank et al. (2007), among
others, show that the wage differential becomes smaller, but does not vanish, once these controls
are included. According to Klein et al. (2013), however, the picture is more nuanced. The
matched employer-employee data for the German manufacturing sector show that exporters pay
a wage premium to high-skilled workers while giving a wage discount to low-skilled workers. A
study by Irarrazabal et al. (2013) on Norwegian manufacturing firms finds that exporters wage
and TFP premia fall substantially once one controls for observable and unobservable worker
characteristics, suggesting that exporters attract more able workers than non-exporters. Recent
theoretical contributions have proposed some mechanisms for why exporters end up employing
more able workers and paying higher wages than non-exporters. Helpman et al. (2010) present
a model where exporters invest more in screeening and as a consequence hire more able workers
who are paid accordingly higher wages. Eckel and Yeaple (2014) consider an environment with
workers heterogeneous in their ability and information frictions, where multi-product exporting
firms are able to use internal labor markets to offer higher compensation to more able workers,
and thus end up paying higher average wages. We contribute to this literature by documenting another dimension along which the compensation structure varies between exporters and
non-exporters, namely, the gender dimension.
The second strand is the large literature on the gender wage gap, reviewed by, for instance,
Blau and Kahn (2000) and OECD (2012). Most closely related to our paper are studies on the
impact of globalization on the gender wage gap. Black and Brainerd (2004) test Beckers model of
discrimination by comparing the difference in residual wage gaps between concentrated relative
to competitive industries that were exposed to comparable increases in import competition.
They find that the gender wage gap narrowed more quickly in the originally more concentrated
industries and interpret this result as supporting the classic predicition of Beckers theory. A
theoretical contribution based on a Melitz-type framework by Ben Yahmed (2013) predicts that
trade integration widens the gender wage gap among skilled workers.8 Our results are line with
these predictions. To the best of our knowledge, this is the first study investigating directly the
relationship between the exporter status and the gender wage gap and doing so using matched
employer-employee data.9
8
In her setting, men and women have the same distribution of ability but differ in their commitment levels.
Firms engage in statistical discrimination which creates the gender wage gap. Firm may choose to invest in a
high technology by incurring an additional cost. Complementarities between commitment/ability and technology
induce exporters (which are the ones choosing to adopt the high technology) to hire more able workers and have
a higher gender wage gap among skilled workers.
9
In a related study, Juhn et al. (2014) examine how trade liberalization affects gender inequality using Mexican
firm-level data. Trade liberalization induces the most productive firms to start exporting and adopt a new tech-
Our work is also related to the recent findings of Goldin (2014) who shows that gender wage
gap is due to employers disproportionately rewarding individuals for laboring long hours and
working particular hours. While her work documents changes in different occupations in this
respect, we focus on differences between different firm types. Moreover, our contribution lies in
using a unique exogenous shock to identify the effects of interest.
Finally, there is a growing number of studies aiming to assess the impact of various gender
equality policies initiatives on male and female labor force participation and earnings in general
(see e.g. Cools et al. (2015), Andresen and Havnes (2014) and Havnes and Mogstad (2011)).
This paper is structured as follows. The next section describes the data and presents some
summary statistics. Section 3 presents the empirical model and the baseline results. Section 4
contains our main analysis where we exploit two exogenous shocks which changed the perceived
work commitment differential between men and women. Section 5 focuses on heterogeneity
among exporters. In Section 6, we address a potential alternative explanation. Section 7 presents
the conclusions.
2
2.1
Data and Basic Facts

Data
We employ a number of different data sets collected by Statistics Norway. We match data on
firms, trade, foreign direct investment and employees for the period 1996-2010. We use firm-level
data from Statistics Norways Capital database, which contains variables such as value added,
output, and employment for an unbalanced panel of all mainland joint-stock companies in the
Norwegian manufacturing sector.10 The manufacturing sector accounts for around 50 percent of
total mainland exports.11 The data set covers around 90% of manufacturing output in Norway.
The industry classification used is the NACE Rev. 2.12 The Capital database is merged with
firm-level data on export and import values based on customs declarations, using a unique firm
identifier. These data make up an unbalanced panel of all yearly export and import values
by 8-digit HS product codes and a destination or source country.13 Finally, we match this
nology that is more modern and requires less physical strength. This improves womens labor market outcomes
in the blue-collar tasks, while leaving them unchanged in the white-collar tasks. The authors use the introduction
of NAFTA to test the model. They find that liberalization is associated with entry of new firms and adoption of
a more modern technology. In addition, firms hire relatively more blue-collar females and increase their share of
the wage bill.
10
Mainland Norway refers to all domestic production activity except from the exploration of crude oil and
natural gas, services activities incidental to oil and gas, transport via pipelines and ocean transport. Statistics
Norways Capital database is described in Raknerud et al. (2004).
11
In 2010, the manufacturing sector accounted for 26% of total exports, 34% of total goods exports and 50%
of total mainland exports, i.e., exports excluding oil, oil-related services and ship- and platform-building. The
top exporting industries within manufacturing were machinery and other equipment n.e.c.; basic metals; basic
chemicals, chemical and mineral products; food products, beverages and tobacco; refined petroleum products.
12
A list of the manufacturing industries can be found in the Appendix. Within the group of manufacturing
industries, the highest share of females among full-time employees can be found in wearing apparel (74.4%),
followed by pharmaceuticals (50.6%) and textiles (43%).
13
In line with findings of studies from other countries, the majority of Norwegian manufacturing firms do not
export. In 1996, only 34.9% were exporting, while in 2005, this number had risen to 38.1%. See Figure 1 in the
merged data set with firm-level data on outward foreign direct investment (FDI). This leaves
us with a comprehensive panel data set on Norwegian manufacturing firms performance and
internationalization, which we then merge with data on their employees.
The main source of employment and wage data for the period 1996 to 2010 is the employee
register (AT) which holds annual records of worked hours and earned wages at the individual
level. Statistics Norway links this register to the tax office database (LTO) to create a correspondence between the wage reported by the employer and those reported to the tax authorities by
the individual. This joint file (ATmLTO) is a much cleaner data set and is therefore used instead
of the AT register. In addition to wages at the person-firm-year level, the database includes the
first and the last date of the employment spell within a given year, the total number of days
worked, the municipalities in which the individual lives and works, and an indicator for full-time
and part-time employment. The ATmLTO data is then merged with time-varying demographic
information about years of education, gender and the number of children, also from Statistics
Norway.
From 2003 onwards, the employment register also contains an occupational classification,
based on the International Standard Classification of Occupations, ISCO-88 (COM). This is a
four-digit code describing the type of job the individual has, ranging from senior officials and
managers to elementary occupations that require no formal education. The codes are grouped
into ten categories.14
The parental leave data come from the Norwegian Labour and Welfare Administration
(NAV). We have annual information on the number of days of paid leave a family is entitled to
after having a child, as well as on the paternal quota, i.e., the share of that leave that is reserved
for the father and cannot be transferred to the mother except under special circumstances. Table 9 in the Appendix provides a overview of the legislation on parental leave in Norway. In
addition, and probably more important, we have data on the actual take-up of the leave. We
know the number of men and women taking some parental leave each year. We also know the
breakdown of how many fathers are taking a certain number days of leave (1-10 days, 11-19
days, 20 days, 21-24 days, 25 days, 26-39 days and more than 40 days.). A week is measured as
the number of workdays, so 20 days of registered leave corresponds to one calendar month.
Finally, we collect data on the regional variation in kindergarten coverage from Statistics
Norway. For each Norwegian municipality, of which there are 439, we have information on the
percentage of children between the ages of 1 and 5 enrolled in kindergarten each year between
2000 and 2010.
Appendix for the share of exporting firms across Norwegian manufacturing industries.
14
These ten categories are: 1. Managers, 2. Professionals, 3. Technicians and associate professionals, 4. Clerks,
5. Service workers and shop and market sales workers, 6. Skilled agricultural and fishery workers, 7. Craft and
related trades workers, 8. Plant and machine operators and assemblers, 9. Elementary occupations, 0. Armed
forces and unspecified. For a more detailed description of the occupational classification, see NOS C 521 Standard
Classification of Occupations, http://www.ssb.no/a/publikasjoner/pdf/nos_c521/nos_c521.pdf.
2.2
Construction of the Sample and Variables
The data set for firms and employees leaves us with a panel covering the population of all
mainland joint-stock manufacturing firms along with their trade, FDI and employees. For each
worker-year combination, we assign to the worker the wage and the firm of the longest employment spell during the year. We restrict our sample to individuals who have worked for at least
three months during a year. We keep only full time employees in the data set. This is done
to avoid biases related to possible part time wage penalties.15 We also restrict our sample to
individuals who are between 19 and 67 years of age and to workers with at least one year of
potential labor market experience. We measure daily wages as the wage earned in that spell,
divided by the number of days worked in that spell, where a spell is defined a worker-firm match.
To remove outliers, we predict wages based on a simple Mincerian regression of log wages on
education, experience and experience squared.16 We then remove observations that lie outside
five times the standard error of the residual. We are left with a sample of 2,713,623 worker-firmyear observations, based on roughly 6,000 firms each year spread across 24 industries (NACE
Rev 2).
We use a host of variables to account for observable worker characteristics. The variable
Education is an individuals total number of years of schooling. We classify a worker as being
College educated if the worker has 14 or more years of education, and Non-college if not. Experience is a measure of actual experience calculated from the Pension register, and gives the
actual number of years a person has been active in the labor market.17 An alternative and often
used measure of labor market experience is a persons potential experience, given by the age of
the worker minus the number of years of education. However, given our focus on gender differences, we choose the actual experience to be our preferrred measure of labor market experience
since women typically have noncontinuous labor market histories and thus potential experience
may be a poor proxy for their actual experience.18 We also include a dummy variable denoted
Children that is equal to one if the worker has at least one child in the year of observation, and
zero otherwise. An alternative demographic characteristic would be workers marital status, but
since cohabitation is very common in Norway also after having children, we believe a persons
private life is better characterized by the Children variable.
To account for geographical aspects of the location of firms and workers, we employ a centrality measure constructed by the Norwegian Institute for Urban and Regional Research (NIBR).
All Norwegian municipalities are grouped into residential and labor market regions. These regions are then split into five categories along the centre-periphery axis, based on the size of the
population, availability of amenities, number of jobs and distance to nearest place categorized
15
It is well known in the gender wage gap literature that part-time workers have lower hourly earnings and
that women are overrepresented among part-time workers, see e.g. Manning and Petrongolo (2008).
16
See below for how the experience and education variables are constructed.
17
The Pension register contains data on incomes dating back to 1967.
18
Note however, that since employees earn pension rights also during parental leave, the registered number of
years in the labor market includes these periods of leave.
as central. The categories are: major cities, medium-sized towns, small towns, rural centres
and periphery. We define a Centrality variable that equals one if an individual is working in a
municipality that belongs to one of the first two categories in the year of observation, and zero
otherwise.
For the firms, Size is a time-varying measure of firm size defined as the log of the number of
employees. We define a firm as being an Exporter if the value of its total exports in the year of
observation exceeds NOK 10,000 (USD 1,100). In addition to whether a firm is an exporter or a
non-exporter, we are also interested in other characteristics of the firm. We measure Profitability
as the log of profits divided by the operating income. MNC is a dummy that equals one if the
firm is registered as having a positive ownership share in a firm located in a foreign country in
the year of observation, and zero otherwise. To proxy for the complexity of a firms operations,
we calculate a variable referred to as Destinations, which is simply the number of destinations
we observe a firm exporting to in a given year.
To address the impact of two exogenous shocks to perceived commitment, we calculate a
paternity leave and a kindergarten variable. The Paternity leave (PL) variable is defined as the
number of men taking 40 or more days (8 or more weeks) of leave in a given year expressed as
a share of all men taking leave in that year. This is done for several reasons. First, we need to
normalize the absolute number of men taking leave in a meaningfull way. Normalizing by the
total number of men or fathers in a given year would introduce issues regarding immigration,
which was substantial during our sample period. Secondly, the 40 days threshold represents a
substantial share of the leave. Our hypothesis that increasing the fathers share of the parental
leave will influence firms perception of female commitment only makes sense if fathers start
taking a significant share of the leave, and we would not expect such effects from fathers taking
a week or two. The choice of a rather high threshold should be seen in the light of the fact that
almost all employees in Norway are entitled to five weeks (25 days) paid vacation every year.
The 40 days threshold is the highest threshold for which we have consistent data throughout
our period. We define the Kindergarten (KG) variable to be equal to the percentage of children
between ages of 1 and 5 enrolled in kindergarten in a given year in the municipality where the
worker lives.
2.3
Some Basic Facts on Women and Exporters
Before we proceed with our core analysis we document some basic facts on women and exporters
in Norway. Table 10 in the Appendix provides an overview of the distribution of workers between
exporters and non-exporters and the average characteristics of the labor force at these two types
of firms. Despite the fact that a minority of firms are exporting, exporters employ around 80
percent of the labor force in manufacturing. We see that exporters on average employ workers
with slightly higher education, slightly more labor market experience and who are slightly older,
but have close to the same number of children. Their employees on average have substantially
higher wages. The average female share in the labor force is approximately the same for exporters
8
and non-exporters.
To have a point of comparison, we estimate the gender wage gap in the Norwegian manufacturing sector by using worker-level data and regressing the log of wages on a female dummy.
We find that the gender wage gap is sizable, and it is higher in manufacturing than in the rest
of the Norwegian economy.19
Table 1: Overall Gender Wage Gap
(1)
-0.248
(0.009)
2713623
0.040
Female
Observations
Adjusted R2
Notes: Estimates are based on the panel of worker-level

data for 1996-2010. Dependent variable: log wage.
Standard errors in parenthesis are clustered on firm.
p < 0.1, p < 0.05, p < 0.01.
Female workers in Norwegian manufacturing on average earn roughly 22 percent less than
men, see Table 1.20 Compared to what Blau and Kahn (2000) report for the US in 1999 (23.5
%), the gender wage gap in Norway is even in the manufacturing sector slightly smaller.
It is well known that exporters are different along a set of performance indicators. It has been
documented for a number of countries, including Norway, that exporters on average pay higher
wages.21 However, it has also been found that international trade and exporters in particular,
contribute to widening of income inequalities both within and between skill groups, see e.g. Klein
et al. (2013). Thus, we procede by documenting some basic facts on the exporter wage premium
and its variation across groups of workers with different educational levels and gender for the
Norwegian manufacturing sector. To calculate the exporter premium, we follow the standard
approach in the literature (see e.g. Bernard et al. (1995)) by regressing the firms average wage
on its exporting status, controlling for firm size and intrustry-year:
ln W agejt = st + Exporterjt + Sizejt + jt ,
(1)
where ln W agejt is the log of the average wage paid by firm j in year t, Exporterjt equals
one if firm j exports in year t, and zero otherwise, and Size depicts firm size, and st denotes
industry-year fixed effects. The results, presented in Table 11 in the Appendix, show that the
estimated coefficient on the exporter status is positive and statistically significant (see Column
(1)). When using the full sample, we find an exporter premium of 5.3% on average.
In Columns (2) and (3) of Table 11, we split the sample into male and female workers and
find that the exporters on average pay higher wages to both sexes, but the exporter premium
19
According to Barth et al. (2013) women in Norway have over the last decade earned between 12.5-15 percent
less than men. Focusing only on the private sector, the gap has averaged 15-17 percent. These gender wage gaps
are estimated based on the entire workforce, including both part-time and full-time employees.
20
The gender wage gap is calculated as 100(e0.250 1) = 22.12.
21
Irarrazabal et al. (2013) find that exporters in the Norwegian manufacturing sector pay higher wages, have
higher productivity, profitability and capital intensity.
is slightly higher for males than females. Education is another worker characteristic regarded
to be important for the extent to which workers gain from globalization. Hence, we split the
sample into non-college and college educated workers (see Columns (4) and (5)). We find that
the exporter premium is substantially higher for the college graduates. Finally, we consider both
gender and education variation in exporter premium, and split the sample into four groups based
on gender and education level (see Columns (6)-(9)). The exporter wage premium is positive
and significant for all groups of workers, but there are substantial differences in the magnitudes.
Within education groups, there seems to be a larger exporter premium for non-college educated
females than males, as low educated males benefit the least from working at an exporter of the
four groups of workers. The exporter wage premium is higher for college than for non-college
educated for both genders, and college educated males are paid a substiantially higher exporter
premium than college educated females.
With this descriptive evidence on gender wage gap and exporter premia as a background we
set out to examine our main question of interest.
Empirical Model and Baseline Results
To shed light on the impact of globalization on gender inequality we examine the relationship
between exporting and the gender wage gap. First, we simply account for observable worker
and firm characteristics. Second, we investigate the role of unobservable worker characteristics
by adding worker or spell fixed effects to our model. In the proceeding section we move on to
exploit two exogenous shocks to cast light on the relationship between exporting and gender
wage gap.
3.1
Exporters and the Gender Wage Gap: Controlling for Observables
The Gender Wage Gap in the Norwegian Manufacturing Sector How large is the
gender wage gap when observable worker and firm characteristics are taken into account? To
answer this question we adopt the approach of the existing literature and estimate a wage
regression of the Mincer (1974) type adding a female dummy. A statistically significant dummy
captures the wage gap between males and females, assuming that the included controls account
for differences in worker productivity. Our model takes the following form:
ln wijst = f emi + Xit + Zjt + st + ijst ,
(2)
where wijst is the daily wage of worker i employed by firm j operating in industry s at time t,
f emi is a dummy for being female, Xit is a vector of observable characteristics of worker i, Zjt
is a vector of observable characteristics of firm j, st is a industry-year fixed effect capturing
any systematic variations in wages across industries, and ijst is an error term. The vector Xit
includes the variables Education, Experience, Children, and the dummy for working in a central
location (Centrality). As is common practice, we also include experience squared (divided by
10
Table 2: Gender Wage Gap
Female
OLS
(1)
-0.224
(0.005)
FE
(2)
-0.218
(0.004)
0.057
(0.001)
0.034
(0.000)
-0.058
(0.001)
-0.026
(0.003)
0.055
(0.006)
0.041
(0.002)
No
Yes
2713623
0.414
0.051
(0.001)
0.032
(0.000)
-0.054
(0.001)
-0.019
(0.002)
0.070
(0.013)
-0.017
(0.004)
Yes
Yes
2713623
0.470
Exporter
Female*Exporter
Education
Experience
Experience2
Children
Centrality
Size
Firm FE
Industry*year
Observations
Adjusted R2
OLS
(3)
-0.272
(0.005)
0.004
(0.006)
0.059
(0.007)
0.057
(0.001)
0.034
(0.000)
-0.058
(0.001)
-0.026
(0.003)
0.055
(0.006)
0.039
(0.002)
No
Yes
2713623
0.415
FE
(4)
-0.263
(0.004)
-0.005
(0.003)
0.054
(0.006)
0.051
(0.001)
0.032
(0.000)
-0.054
(0.001)
-0.019
(0.002)
0.070
(0.013)
-0.017
(0.004)
Yes
Yes
2713623
0.470
Notes: Estimates are based on the panel of worker-level data for 1996-2010.
Dependent variable: log wage. Standard errors in parenthesis are clustered
on firm. p < 0.1, p < 0.05, p < 0.01.
100). Zjt for now includes the variable measuring firm size (Size). We refer to Section 2.2 for
details on the variables and their construction. Standard errors are clustered at the firm level
to account for group correlation across workers within firms.22
The results, given in Column (1) of Table 2, show that after controlling for education,
experience and children, women earn roughly 20% less than men. The coefficients on the other
variables have the expected signs, and their magnitudes are in line with the existing literature.
Both experience and education are positively correlated with wages, while having children has a
negative effect. Workers working for larger and more centrally located firms tend to earn more.
The estimate of the gender wage gap is merely about two percentage points lower than
the overall gender wage gap in the manufacturing sector without controlling for worker or firm
characteristics. Hence, our estimates tell us that only around 10 percent of the gender wage gap
in the Norwegian manufacturing can be explained by observable worker characteristics. However,
it is important to bear in mind that we are limiting our analysis to one sector of the economy
and that a substantial part of the gender wage gap in Norway is explained by women primarily
working in different sectors than men do (see e.g. Barth et al. (2013)).
To understand whether the result on the gender wage gap is driven mainly by between-firm
or within-firm variation, we estimate the model above adding firm fixed effects. The results
22
All of our results are robust to clustering at the level of individual workers.
11
from this regression, presented in Column (2) of Table 2, show that the within-firm gender gap
is only slightly lower than the overall gender gap. This tells us that the aggregate gender wage
gap stems mostly from female and male workers being paid different wages within a firm rather
than being due to differences across firms.
Exporting and the Gender Wage Gap We now examine whether there is a difference in
the gender wage gap between exporters and non-exporters. We do so by adding a dummy for
exporting firms, as well as an interaction term between a firms export status and the female
dummy. The latter variable picks up any differential in the gender wage gap between exporting
and non-exporting firms, controlling for the observable characteristics of the worker and the
firm. Hence, the regression we estimate is given by:
ln wijst = f emi + f emi Exporterjt + Xit + Zjt + st + ijst ,
(3)
where the vector of observable characteristics of firm j, Zjt , now includes also the export status
of the firm. The results from OLS and firm fixed effects regressions are given in Columns (3)
and (4) of Table 2, respectively. Once we control for worker characteristics as well as firm
size, we observe that the exporter premium is no longer statistically significantly different from
zero.23 The results further show that the gender wage gap is still statistically significant and
economically large when we account not only for observable worker and firm characteristics, but
also for the exporting status. This is true both for the OLS and firm fixed effects regressions.
Most interesting from our point of view is the fact that female workers seem to get a positive
and significant exporter wage premium. The interaction term between being female and working
for an exporting firm is positive and statistically significant. The coefficient is almost the same
regardless of whether or not we include firm fixed effects. We find that in general females are
paid 24 percent less, while women working at exporting firms are paid only around 19 percent
less than their male peers. Thus working at an exporter seems to close the gender wage gap by
around a fifth.
Subsample Analysis according to Education and Occupation In the exploratory firmlevel regressions (see Table 11 in the Appendix), we found large differences between education
groups, with the college graduates enjoying a larger exporter premium than less educated workers. To examine this issue further, we split the sample into workers with and without college
education and estimate equation (3) separately on each sub-sample. The results are given in
Table 12 in the Appendix. They show that college educated workers in general get an exporter
premium, even after we control for their observable characteristics and the size of the firm. This
is, however, not the case for the workers without college education, where the exporter wage
premium vanishes once we control for their observable characteristics and the size of the firm.
23
If we exclude firm size from the set of control variables, the exporter premium is again positive and significant
in the OLS estimation. This is not surprising given that most exporters are large firms.
12
The relative size of the two groups explain why the exporter premium disappears when we pool
all workers: the non-college group constitues roughly 80% of the sample (Table 2). However, in
both education groups, females get a sizable exporter premium. This result holds both across
firms (Columns (1) and (3)) and within firms (Columns (2) and (4) where we add firm fixed
effects). The exporter female premium is larger for workers without college education. As for
the a general gender wage gap, this is found for both groups of education, but our estimates
suggest that the gap is larger among workers without college education.
For the last part of our sample period, from 2003 onwards, we also know the occupation
of the worker. While education may be suggestive of the type of work and tasks that an
employee carries out, occupation may be even more informative. We proceed by splitting the
sample into eight occupational categories: Managers (Mng), Professionals (Prof), Technicians
and associate professionals (Tech), Clerks, Service workers and shop and market sales workers
(Service), Craft and related trades workers (Craft), Plant and machine operators and assemblers
(Mchn), and Elementary occupations (Elem), leaving out two categories that are not relevant for
manufacturings firms, namely Armed forces and unspecified and Skilled agricultural and fishery
workers. The results are reported in Table 13 in the Appendix.
We see that across all occupations, there is a statistically significant gender wage gap. This
gap is smallest for Professionals, and largest for Service workers. Exporters pay a wage premium
to Managers, Technicians and associate professionals, and Service workers, while Clerks as well
as Plant and Machine operators and assemblers actually earn relatively less at exporters. These
results seem to support the findings of Klein et al. (2013). In five of eight occupational categories,
exporters pay an extra premium to females. The only statistically significant negative exporter
effect on the gender wage gap is found for the group of Professionals. We also repeat this exercise
adding firm fixed effects, but we do not report the results to save space. Again in five occupations
we find that the gender wage gap is smaller in exporting firms. The result suggesting a larger
gender wage gap in exporting firms among Professionals ceases to be statistically significant.
Allowing for Differential Returns to Worker Characteristics It might be the case
that the exporters and non-exporters differ not only in terms of the gender wage gap but also in
terms of the returns they offer to other worker characteristics, such as, education and experience.
Therefore, we extend our specification to to allow for differential returns to all variables. We
proceed first by allowing for returns to vary between exporting and non-exporting firms (Column
(1)), then between females and males (Column (2)), and finally between all four subgroups
(Column (3)).24 The results are presented in Table 14 in the Appendix. Our results indicate
that the return to observable characteristics such as education experience and children vary both
across gender and between exporters and non-exporters. But most importantly, we find that
the general exporter premium to females remains also after these differential returns have been
accounted for.
24
Our reference group is males in non-exporting firms.
13
3.2
Exporters and the Gender Wage Gap: Controlling for Unobservables
Our analysis so far has revealed persistent differences in the gender wage gap between exporters
and non-exporters when controlling for observable characteristics. What if this is because the
females working for exporters are different in ways the econometrician cannot observe? May the
lower gender wage gap we observe in exporting firms simply be an artifact that can be explained
by exporters employing different women than non-exporters? Such differences in firms labor
force could be explained by some kind of sorting, where women with different abilities and
ambitions seek to be employed by exporters, or due to exporters being better at screening and
thus employing women with higher productivity than non-exporters. To explore this issue we
need to control for unobservable characteristics of the workers which affect their productivity.
The most straightforward way of doing so is by adding worker fixed effects to the regressions.
These fixed effects will account for all time-invariant heterogeneity of the workers, including
their gender.25 This means that we will not be able to identify the overall gender wage gap
using this methodology. However, our interaction of interest between being female and working
for an exporter is still identified in this framework. The coefficient on this variable will tell us
how large the differential gender wage gap between exporters and non-exporters is, conditional
on observable, time-varying characteristics of the worker and the firm, as well as unobservable
time-invariant characteristics of the worker. If we believe that ability and ambition do not
change over time, they will now be accounted for in the analysis. We proceed by estimating a
model based on
ln wijst = f emi Exporterjt + Xit + Zjt + st + i + ijst ,
(4)
where Xit is again a vector of time-varying characteristics of the worker, Zjt is a vector of timevarying characteristics of the firm including its exporter status, st is the industry-year fixed
effect, and i is the worker fixed effect.26 The results from this regression are given in Table 3,
for the whole sample (Column (1)), as well as for the two education groups separately (Columns
(2)-(3)).
These results tell a very different story. The coefficient on our interaction variable F emale
Exporter is statistically significant and negative for college graduates. In the pooled sample and
the subsample of workers without college education, it also bears a negative sign but it is not
statistically significant. In other words, when we account for unobservable heterogeneity among
the workers, females no longer earn a gender-specific exporter premium. This suggests that the
observable characteristics that we included in our previuos regressions did not capture all the
relevant aspects of a workers productivity and that exporters hire more able workers, in terms
of unobservable characteristics. Taking these unobservable characteristics into account, there is
a wage penalty for college educated women working in exporting firms of about 3 percentage
25
No workers in our sample switch gender.

Recall that the Centrality dummy is defined based on where the worker lives. Since workers move during a
spell, we are able to include this variable in the regressions.
26
14
Table 3: Worker and Spell Fixed Effects
Exporter
Female*Exporter
Education
Experience
Experience2
Children
Centrality
Size
Worker FE
Spell FE
Industry*year
Observations
Adjusted R2
All
(1)
0.005
(0.003)
-0.007
(0.004)
0.066
(0.002)
0.012
(0.005)
-0.059
(0.001)
-0.057
(0.004)
0.018
(0.004)
0.027
(0.002)
Yes
No
Yes
2713623
0.714
Worker FE
Non
(2)
0.005
(0.003)
-0.003
(0.005)
0.076
(0.003)
0.027
(0.006)
-0.056
(0.001)
-0.074
(0.004)
0.014
(0.005)
0.030
(0.002)
Yes
No
Yes
2160012
0.656
Coll
(3)
0.009
(0.006)
-0.029
(0.008)
0.089
(0.006)
0.013
(0.006)
-0.067
(0.002)
-0.030
(0.007)
0.020
(0.005)
0.016
(0.002)
Yes
No
Yes
553611
0.781
All
(4)
0.003
(0.004)
-0.010
(0.005)
0.063
(0.002)
0.020
(0.008)
-0.053
(0.001)
-0.066
(0.006)
0.001
(0.007)
0.045
(0.006)
No
Yes
Yes
2713623
0.750
Spell FE
Non
(5)
0.004
(0.004)
-0.007
(0.005)
0.075
(0.003)
0.041
(0.010)
-0.050
(0.001)
-0.083
(0.006)
-0.004
(0.008)
0.044
(0.006)
No
Yes
Yes
2160012
0.696
Coll
(6)
0.008
(0.006)
-0.032
(0.010)
0.063
(0.008)
0.013
(0.011)
-0.064
(0.002)
-0.035
(0.009)
0.010
(0.007)
0.038
(0.008)
No
Yes
Yes
553611
0.813
Notes: Estimates are based on the panel of worker-level data for 1996-2010. Dependent variable:
log wage. Standard errors in parenthesis are clustered on firm. p < 0.1, p < 0.05, p < 0.01.
points. Note that the identifying variation in this model comes from firms changing their export
status and from workers moving between exporters and non-exporters. In additional regressions,
not reported to save space, we find that both sources of variation are responsible for our findings.
Next, we rely on spell fixed effects rather than worker fixed effects to investigate the role
of unobservables. A spell is defined as a unique worker-firm combination. This strategy uses
variation coming only from firms changing their export status, holding all worker- and matchspecific time-invariant effects constant. In this way, we circumvent the problem of endogenous
mobility and are able to account for firm-specific unobservable characteristics at the same time.27
The results, presented in the last three columns of Table 3, are very similar in sign and magnitude
to the results with worker fixed effects. A female wage penalty of about 3.3 percentage points is
found among workers with college education. A smaller, but still weakly statistically significant,
effect of 1 percentage point is present in the full sample.
27
We could also use firm and worker fixed effects jointly to take out firm effects, but that would would still
leave us with the problem of endogenous mobility.
15
Women, Commitment and the Gender Wage Gap: Exploiting

Exogenous Shocks
As argued above, men are typically believed to be more committed employees than women. Over
the last decades Norwegian policy makers have introduced a number of policy measures intended
to promote equal rights and emancipation. Among them are a set of subsequent paternity leave
reforms and a roll-out of subsidized day care for children. Doing so, policymakers have aimed to
influence the relationship between partners - mothers and fathers, between parents and children,
but also between employers and employees. We argue that, as a consequence, social attitudes
have changed and so did the perceived commitment gap between the genders. If exporters are
more dependent on employees commitment and flexibility, our hypothesis is that this change
in perceived commitment will matter relatively more for these firms than for other firms. As
a result, we would expect the difference in relative gender wage gap between exporters and
non-exporters to be reduced. Below we proceed by exploiting the two exogenous shocks to
emanipation - paternity leave and increased kindergarten coverage - and examine the impact of
these on the gender wage gap differential between exporters and non-exporters. .
4.1
Paternity Leave Reforms
The first exogenous shock we consider consisted of legislative changes that have gradually increased the number of weeks of the parental leave that are available only to the childs father,
the so called paternal quota. Unless there are special circumstances, the paternal quota weeks
are lost if not taken by the father. A quota of 4 weeks was first introduced in 1993. It was
extended to 5 and 6 weeks in 2005 and 2006, respectively. In 2009, the number of weeks was
further extended to 10. Table 9 in the Appendix provides an overview of the subsequent reforms.
These legislative changes have resulted in a huge increase in the fraction of fathers taking at
least 8 weeks of leave. This share went up from 8.4% in 1996 (the first year of our sample), to
10.8% in 2003 and 40.6% in 2010 (the last year of our sample).
We argue that it is likely that the change in fathers behaviour has had an impact on the
perceived commitment gap between men and women. Our presumption is supported by timeuse surveys of parents and the time they spend on domestic tasks (see Kitterd (2012) ). The
surveys show that over the last couple of decades mothers in Norway have consistently reduced
the time they spend on childcare and household chores, while the opposite is true for fathers.
In our empirical analysis, we proxy for the change in the perceived commitment gap between
men and women, using the fraction of fathers taking at least 8 weeks of leave in a given year.
As argued in Section 2.2, we choose 8 weeks as our cutoff because a two-month long absence
from work is substantial and disruptive enough to affect the employers perceptions. In line
with our hypothesis that a change in the perceived commitment gap should affect exporters
relatively more, our variable of interest is a triple interaction between the female dummy, the
exporter dummy and the parental leave variable. It captures us how the gender wage gap
16
differential between exporters and non-exporters evolved with the change in social attitudes.
The specification also includes the other components of the triple interaction separately.
The estimation results, presented in Table 4, confirm our priors. We find that the gender
wage gap is higher in exporting firms than in non-exporters, and more importantly that the
difference between the two narrows down with the change in social attitudes. All of the estimated
coefficients are statistically significant, with all but one being significant at the one percent level.
This is true for the full sample as well as for subsample of college educated workers and workers
without college education. As one would expect, the results are stronger when we focus our
attention to workers in their reproductive years, i.e., those under 45 years of age (see the last
three columns of the table). They are also stronger for college graduates.
Table 4: Paternity leave and the Gender Wage Gap
Exporter
Female*Exporter
Female*Exporter*PL
Exporter*PL
Education
Experience
Experience2
Children
Centrality
Size
Spell FE
Female*year
Industry*year
Observations
Adjusted R2
All
(1)
-0.014
(0.005)
-0.029
(0.007)
0.185
(0.036)
0.136
(0.029)
0.063
(0.002)
0.021
(0.008)
-0.054
(0.001)
-0.067
(0.006)
0.001
(0.007)
0.045
(0.006)
Yes
Yes
Yes
2713623
0.750
All Workers
Non
Coll
(2)
(3)
-0.010
-0.014
(0.005)
(0.008)
-0.027 -0.047
(0.007)
(0.014)
0.196
0.125
(0.039)
(0.067)
0.108
0.164
(0.027)
(0.042)
0.075
0.063
(0.003)
(0.008)
0.042
0.014
(0.010)
(0.011)
-0.051
-0.065
(0.001)
(0.002)
-0.084
-0.035
(0.006)
(0.009)
-0.004
0.011
(0.008)
(0.007)
0.045
0.038
(0.006)
(0.008)
Yes
Yes
Yes
Yes
Yes
Yes
2160012
553611
0.696
0.813
Workers < 45 years old

All
Non
Coll
(4)
(5)
(6)
-0.009
-0.005
-0.013
(0.006)
(0.005)
(0.008)
-0.039 -0.034
-0.063
(0.010)
(0.010)
(0.018)
0.251
0.254
0.219
(0.052)
(0.058)
(0.090)
0.104
0.072
0.135
(0.032)
(0.032)
(0.049)
0.069
0.083
0.062
(0.003)
(0.003)
(0.008)
0.009
0.030
-0.009
(0.011)
(0.012)
(0.013)
-0.064
-0.074
-0.047
(0.004)
(0.005)
(0.005)
-0.076
-0.096
-0.043
(0.006)
(0.006)
(0.009)
-0.000
-0.005
0.008
(0.008)
(0.008)
(0.008)
0.045
0.045
0.031
(0.006)
(0.006)
(0.008)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
1621583
1267545
354038
0.733
0.688
0.787
Notes: Estimates are based on the panel of worker-level data for 1996-2010. Dependent variable: log wage.
PL is year specific and is defined as the share of men on paternity leave taking at least 8 weeks of leave.
Standard errors in parenthesis are clustered on firm. p < 0.1, p < 0.05, p < 0.01.
The magnitude of the estimated effects is economically meaningful. Let us consider the
estimates from the last column, i.e., those pertaining college educated workers under 45 years of
age. The change in the share of fathers taking at least two months of paternity leave between
1996 and 2008 results in the gender wage gap in exporters going from being on average 4.5
percent higher than in other firms to being only 3.3 percent higher. For the subsample of
workers without college education, the corresponding gap of 1.3 percent is driven all the way
17
down to zero.
4.2
Kindergarten Roll-out
Now we move on to exploit an alternative shock in our identification strategy. In 2003, the
Norwegian government initiated a substantial expansion of kindergarten coverage. The goal was
to offer all children a high quality, low price place in a public or private kindergarten. Possibly
just as important, the roll out of kindergarten places also meant much more reliable as well as
full time day care. The initiative resulted in a massive change: 9 of 10 children between the
ages of 1 and 6 attended kindergarten in 2009, as compared to 54% in 1996 and 63% in 2001.
The increase in kindergarten coverage proceeded at different speeds in different regions of the
country.
In our estimation, we examine whether the gender wage gap differential between exporters
and non-exporters was affected by the increase in availability of kindergarten places in a given
location. The measure of interest is the share of children between the ages of 1 and 5 enrolled
in kindergarten. The logic behind this approach is that greater availability of reliable child care
is associated with fewer disruptions to mothers working hours and thus allows them to exhibit
greater commitment to their jobs. Thus the availability of kindergartens should be negatively
correlated with the perceived differences in work commitment between genders, which matters
relatively more for commitment-dependent firms. As the speed of the kindergarten roll-out may
have been related to locational characteristics, we control for county-year fixed effects.28
Table 5 shows that our results on the impact of the kindergarten reform are consistent with
those found for the paternity leave (see Table 4). We find that the gender wage gap differential
between exporters and non-exporters went down as a larger share of children of the relevant
age was enrolled in kindergarten in a given location in a given year. The coefficients of interest
are statistically significant in all specifications. As expected, the gender wage gap differential
between exporters and non-exporters is higher for women in their reproductive years (i.e., those
under 45).
The estimates are also economically meaningful. Let us again focus on the last column of the
table pertaining to college educated workers under 45 years of age. The estimated coefficients
suggest that an increase in the average kindergarten coverage from that observed in 2000 (the
first year for which such data are available) to that observed in 2009 would change the gender
wage gap differential between exporters and non-exporters from about 5 percent to zero. In
other words, it would eliminate the wage penalty faced by college educated females working for
exporting firms.
28
There are 439 municipalities and 19 counties in Norway.
18
Table 5: Kindergarten and the Gender Wage Gap
Exporter
Female*Exporter
Female*Exporter*KG
Exporter*KG
Female*KG
KG
Education
Experience
Experience2
Children
Centrality
Size
Spell FE
County*year
Industry*year
Observations
Adjusted R2
All
(1)
-0.053
(0.016)
-0.150
(0.028)
0.002
(0.000)
-0.002
(0.000)
0.001
(0.000)
-0.001
(0.000)
0.073
(0.003)
0.033
(0.008)
-0.052
(0.001)
-0.068
(0.007)
0.010
(0.006)
0.047
(0.009)
Yes
Yes
Yes
1896790
0.757
All Workers
Non
(2)
-0.042
(0.015)
-0.152
(0.029)
0.002
(0.000)
-0.003
(0.000)
0.001
(0.000)
-0.000
(0.000)
0.088
(0.004)
0.041
(0.009)
-0.049
(0.001)
-0.087
(0.007)
0.010
(0.008)
0.045
(0.007)
Yes
Yes
Yes
1485425
0.700
Coll
(3)
-0.052
(0.035)
-0.167
(0.070)
0.002
(0.001)
-0.002
(0.001)
0.001
(0.000)
-0.000
(0.000)
0.057
(0.009)
0.029
(0.015)
-0.060
(0.002)
-0.034
(0.012)
0.005
(0.010)
0.042
(0.012)
Yes
Yes
Yes
411365
0.816

All
Non
Coll
(4)
(5)
(6)
-0.029
-0.015
-0.037
(0.018)
(0.018)
(0.037)
-0.189 -0.197 -0.178
(0.040)
(0.042)
(0.085)
0.002
0.003
0.002
(0.001)
(0.001)
(0.001)
-0.003
-0.003 -0.003
(0.000)
(0.001)
(0.001)
0.000
0.000
0.001
(0.000)
(0.000)
(0.000)
-0.000
-0.000
0.000
(0.000)
(0.000)
(0.001)
0.080
0.097
0.057
(0.003)
(0.004)
(0.009)
0.020
0.028
0.005
(0.011)
(0.011)
(0.019)
-0.054
-0.064 -0.033
(0.004)
(0.005)
(0.006)
-0.076
-0.096 -0.041
(0.007)
(0.007)
(0.012)
0.006
0.007
-0.000
(0.008)
(0.010)
(0.012)
0.045
0.043
0.038
(0.008)
(0.007)
(0.011)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
1099014
842514
256500
0.736
0.689
0.786
Notes: Estimates are based on the panel of worker-level data for 2000-2010. Dependent variable: log
wage. KG is year and municipality specific and is defined as the share of children between the ages of
1 and 5 enrolled in kindergarten. Norway has 19 counties and 439 municipalities.
19
Exporter Heterogeneity and the Gender Wage Gap
So far our analysis has focused on differential gender wage gaps beween exporters and nonexporters. However, exporters differ substantially in their size, organization and technology.
Now we set out to exploit the heterogeneity among exporters. It has been documented that
firms exporting to a large number of locations tend to be larger and more productive (Bernard
et al. (2011), Eaton et al. (2011)). It is also likely that these firms have even higher expectations
vis a vis their employees than other exporters, which may translate into them having an even
larger gender wage gap.
Table 6: Number of Export Destinations
1-3 Destinations
4-10 Destinations
>10 Destinations
Female*1-3 Destinations
Female*4-10 Destinations
Female*>10 Destinations
Education
Experience
Experience2
Children
Centrality
Size
Spell FE
Industry*year
Observations
Adjusted R2
All
(1)
0.004
(0.003)
0.004
(0.006)
0.007
(0.007)
-0.010
(0.005)
-0.012
(0.007)
-0.024
(0.009)
0.063
(0.002)
0.020
(0.008)
-0.053
(0.001)
-0.066
(0.006)
0.001
(0.007)
0.045
(0.006)
Yes
Yes
2713623
0.750
Non-college
(2)
0.005
(0.003)
0.003
(0.006)
0.006
(0.007)
-0.009
(0.005)
-0.007
(0.007)
-0.020
(0.009)
0.075
(0.003)
0.041
(0.010)
-0.050
(0.001)
-0.083
(0.006)
-0.004
(0.008)
0.044
(0.006)
Yes
Yes
2160012
0.696
College
(3)
0.005
(0.006)
0.015
(0.009)
0.016
(0.011)
-0.025
(0.011)
-0.039
(0.014)
-0.043
(0.017)
0.063
(0.008)
0.013
(0.011)
-0.064
(0.002)
-0.035
(0.009)
0.010
(0.007)
0.038
(0.008)
Yes
Yes
553611
0.813
on firm. p < 0.1, p < 0.05, p < 0.01.
To examine this issue we split the exporting firms into three bins: those exporting to between
1 and 3 destinations, 4-10 destinations and those serving more than 10 destinations. Then we
allow the gender wage gap to be different across bins and estimate the model given by (4) using
spell fixed effects. As expected, we find that the coefficient on theF emale Exporter interaction
becomes more negative as the number of destinations increases (see Table 6). This means that
the gender wage gap is higher in exporters serving a larger number of markets. While the
20
difference between estimated coefficients of interest is statistically significant in the full sample
and the subsample of workers without college education, this is not true for college educated
workers.
Addressing an Alternative Explanation
Finally, we turn to a possible alternative interpretation for our baseline findings of a higher gender
wage gap among exporting firms. Beckers (1957) theory of taste-based discrimination predicts
that only very profitable firms can afford to discriminate. As exporters are more profitable than
other firms, they may be better positioned to engage in costly discrimination against women.
In other words, Beckers theory would lead to the same prediction of a higher gender wage gap
among exporters, but its interpretation would be different.
Table 7: Multinationals, Profitability and Size
Exporter
Female*Exporter
MNC
Female*MNC
Size
Female*Size
Profitability
Female*Profitability
Education
Experience
Experience2
Children
Centrality
Spell FE
Industry*year
Observations
Adjusted R2
All
(1)
0.003
(0.004)
-0.008
(0.005)
-0.016
(0.008)
-0.001
(0.006)
0.051
(0.006)
-0.023
(0.006)
0.006
(0.001)
-0.001
(0.001)
0.063
(0.002)
0.020
(0.008)
-0.053
(0.001)
-0.066
(0.006)
0.001
(0.007)
Yes
Yes
2713623
0.750
Non-college
(2)
0.003
(0.004)
-0.006
(0.005)
-0.012
(0.007)
0.002
(0.007)
0.050
(0.006)
-0.022
(0.006)
0.006
(0.001)
-0.002
(0.001)
0.075
(0.003)
0.041
(0.010)
-0.050
(0.001)
-0.083
(0.006)
-0.004
(0.008)
Yes
Yes
2160012
0.696
College
(3)
0.007
(0.006)
-0.030
(0.011)
-0.017
(0.010)
-0.002
(0.008)
0.042
(0.009)
-0.018
(0.010)
0.006
(0.001)
0.001
(0.001)
0.063
(0.008)
0.013
(0.012)
-0.064
(0.002)
-0.035
(0.009)
0.011
(0.007)
Yes
Yes
553611
0.813
Dependent variable: log wage. Dummy for missing profitability included.
Standard errors in parenthesis are clustered on firm.
p < 0.1, p < 0.05, p < 0.01.
21
To address this alternative explanation we augment our specification with spell fixed effects
from Section 3.2 by allowing the gender wage gap to differ with firm profitability.29 We also
allow it to differ with firm size (which in the Melitz (2003) framework is positively correlated
with profitability) and its multinational status (again because it is a proxy for profitability). As
evident from Table 7, doing so does not affect our main result suggesting that the gender wage
gap is higher in exporting firms among college educated workers.
We also augment our exercises relying on the exogenous shocks by controlling for profitability
and its interaction with the female dummy. As evident from Tables 15 and 16 in the Appendix,
our results are robust to this change.
Conclusions
While the impact of globalization on income inequality has received a lot of attention, little is
yet known about its effect on the gender wage gap. This study argues that there is a systematic
difference between the gender wage gap in exporting firms and non-exporters. By the virtue of
being exposed to higher competition, exporters require greater commitment and flexibility from
their employees. If commitment is not easily observable and women are perceived to be less
committed workers than men, exporters will statistically discriminate relatively more against
female employees and will exhibit a higher gender wage gap than non-exporters. Moreover, we
should expect a decrease in the perceived commitment gap between men and women to narrow
the differential gender wage gap between these two types of firms.
We test this hypothesis using the matched employer-employee data from the Norwegian
manufacturing sector for the 1996-2010 period. Our identification strategy relies on an exogenous
shock, the legislative changes that increased the number of weeks of the parental leave that are
available only to the childs father. We argue that these changes have narrowed the perceived
commitment gap between the genders. We show that the initially higher gender wage gap
observed in exporting firms (relative to non-exporters) has gone down after the shock. This
effect was particularly pronounced for college graduates and for those in their reproductive
years.We also show that another exogenous shock an increase in the kindergarten coverage
has lead to a similar effect, and served as to close the gender wage gap between exporters and
non-exporters.
29
The Profitability variable is missing for approximately 22% of the sample. Instead of dropping these observations, we add a dummy for missing Profitability in all the specifications where Profitability is included.
22
References
Andresen, M. E. and T. Havnes (2014). Money for nothing? Universal child care and maternal
employment. Journal of Public Economics.
Barth, E., I. Hardoy, P. Schoene, and K. M. Oestbakken (2013). Lnnsforskjeller mellom kvinner
og menn. Hva skjedde p 2000-tallet? ISF Report 2013:7, Institute for Social Research.
Becker, G. S. (1957). The Economics of Discrimination. Chicago: University of Chicago Press.
Ben Yahmed, S. (2013). Gender wage gaps across skills and trade openness. Working paper.
Bernard, A. B. and J. B. Jensen (1999). Exceptional exporter performance: Cause, effect, or
both? Journal of International Economics 47 (1), 125.
Bernard, A. B., J. B. Jensen, and R. Z. Lawrence (1995). Exporters, jobs, and wages in US
manufacturing: 1976-1987. Brookings Papers on Economic Activity. Microeconomics 1995,
67119.
Bernard, A. B., S. J. Redding, and P. K. Schott (2011). Multiproduct firms and trade liberalization. The Quarterly Journal of Economics 126 (3), 12711318.
Black, S. E. and E. Brainerd (2004). Importing equality? The impact of globalization on gender
discrimination. Industrial and Labor Relations Review 57 (4), 540559.
Blau, F. D. and L. M. Kahn (2000). Gender differences in pay. Journal of Economic Perspectives 14 (4), 7599.
Cools, S., J. H. Fiva, and L. J. Kirkeben (2015). Causal effects of paternity leave on children
and parents. Scandinavian Journal of Economics forthcoming.
Cuddy, Amy, S. F. and P. Glick (2004). When professionals become mothers, warmth does not
cut the ice. Journal of Social Issues 60 (4), 701718.
Eaton, J., S. Kortum, and F. Kramarz (2011). An anatomy of international trade: Evidence
from french firms. Econometrica 79 (5), 14531498.
Gislason, I. (2007). Parental leave in Iceland: Bringing the fathers in. Technical report, Centre
for Gender Equality, Ministry of Social Affairs.
Goldin, C. (2014). A grand gender convergence: Its last chapter. American Economic Review 104 (4), 10911119.
Havnes, T. and M. Mogstad (2011). Women and children first? Labor market effects of universal
child care for toddlers. Technical Report 11-12.
Helpman, E., O. Itskhoki, and S. Redding (2010). Inequality and Unemployment in a Global
Economy. Econometrica 78 (4), 12391283.
23
Irarrazabal, A., A. Moxnes, and K. H. Ulltveit-Moe (2013). Heterogeneous firms or heterogeneous

workers? Implications for the exporter premium and the impact of labor reallocation on
productivity. Review of Economics and Statistics 95 (3), 839849.
Juhn, C., G. Ujhelyi, and C. Villegas-Sanchez (2014). Men, women and machines: How trade
impacts gender inequality. Journal of Development Economics 106, 179193.
Kitterd, R. H. (2012). Foreldrenes tidsbruk: Fedre deltar med i husarbeid og omsorg. Statistics
Norway Samfunnsspeilet (4).
Klein, M. W., C. Moser, and D. M. Urban (2013). Exporting, skills and wage inequality. Labour
Economics 25, 7685.
Manning, A. and B. Petrongolo (2008). The part-time pay penalty for women in Britain. The
Economic Journal 118, F28F51.
Melitz, M. J. (2003). The Impact of Trade on Intra-Industry Reallocations and Aggregate
Industry Productivity. Econometrica 71 (6), 16951725.
Mincer, J. (1974). Schooling, Experience, and Earnings. Columbia University Press.
OECD (2012). Closing the Gender Gap: Act Now. OECD Publishing.
Raknerud, A., D. Rnningen, and T. Skjerpen (2004). Documentation of the capital database.
Statistics Norway Reports 2004/16, Statistics Norway.
Schank, T., C. Schnabel, and J. Wagner (2007).
Do exporters really pay higher wages?
First evidence from German linked employeremployee data. Journal of International Economics 72 (1), 5274.
24
A
A.1
Appendix
Manufacturing Industries
Table 8: NACE Rev. 2
NACE code
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Industry name
Manufacture of food products
Manufacture of beverages
Manufacture of tobacco products
Manufacture of textiles
Manufacture of wearing apparel
Manufacture of leather and related products
Manufacture of wood and products of wood and cork, except furniture;
manufacture of articles of straw and plaiting materials, except furniture
Manufacture of paper and paper products
Printing and reproduction of recorded media
Manufacture of coke and refined petroleum products
Manufacture of chemicals and chemical products
Manufacture of basic pharmaceutical products and pharmaceutical preparations
Manufacture of rubber and plastic products
Manufacture of non-metallic mineral products
Manufacture of basic metals
Manufacture of fabricated metal products, except machinery and equipment
Manufacture of computer, electronic and optical products
Manufacture of electrical equipment
Manufacture of machinery and equipment n.e.c.
Manufacture of motor vehicles, trailers and semi-trailers
Manufacture of other transport equipment
Manufacture of furniture
Other manufacturing
Repair and installation of machinery and equipment
25
.2
.4
Freq
.6
.8
Figure 1: Share of exporters across industries
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
A.2
Additional Tables
Table 9: Parental leave reforms in Norway
July, 1 1977
May 1, 1987
July 1, 1988
April 1, 1989
May 1, 1990
July 1, 1991
April 1, 1992
April 1, 1993
July 1, 2005
July 1, 2006
July 1, 2009
July 1, 2011
Parental leave
(weeks)
18
20
22
24(30)
28(35)
32(40)
35(44.4)
42(52)
43(53)
44(54)
46(56)
47(57)
Compensation
Rate
100%
100%
100%
100%(80%)
100%(80%)
100%(80%)
100%(80%)
100%(80%)
100%(80%)
100%(80%)
100%(80%)
100%(80%)
26
Maternal quota
(weeks)
6
6
6
6
6
8
8
9
9
9
9
9
Paternal quota
(weeks)
0
0
0
0
0
0
0
4
5
6
10
12
Table 10: Descriptive Statistics

Average Wage
Average Education
Average Experience
Average Age
Children share
Average number of children
Centrality
Total share of labor force
Average Share of Females in the labor force
Non-exporters
805.69
12.26
21.06
40.09
0.80
1.84
0.47
0.19
0.20
Exporters
989.27
12.93
22.64
41.60
0.81
1.81
0.53
0.81
0.21
Notes: All numbers apart from average wages are based on the panel of worker-level data for 1996-2010.
Average wages are for 2001. An exporters has by definition exports above NOK 10,000 (USD 1,100).
Table 11: Exporter premia

All
Non
Coll
Non
M
Exporter
Non
F
Coll
M
Coll
F
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
0.052 0.052 0.043 0.035 0.087 0.033 0.037 0.097 0.052
(0.005) (0.005) (0.006) (0.004) (0.007) (0.004) (0.007) (0.007) (0.011)
Size
0.032 0.036 0.050 0.032 0.049 0.035 0.048 0.053 0.050

(0.002) (0.002) (0.003) (0.002) (0.003) (0.002) (0.003) (0.003) (0.004)
Industry*year
Observations
Adjusted R2
Yes
89779
0.322
Yes
87386
0.309
Yes
54551
0.304
Yes
86830
0.314
Yes
51700
0.262
Yes
83907
0.301
Yes
50885
0.289
Yes
47456
0.266
Yes
22281
0.261
Notes: Estimates are based on the panel of firm-level data for 1996-2010. Dependent variable: log
wage. Standard errors in parenthesis are clustered on firm. p < 0.1, p < 0.05, p < 0.01.
M stands for Males, F for Females, Coll for College educated, Non for Non-College educated.
27
Table 12: Gender Wage Gap Education Split
Female
Exporter
Female*Exporter
Education
Experience
Experience2
Children
Centrality
Size
Firm FE
Industry*year
Observations
Adjusted R2
Non-college
(1)
(2)
-0.278 -0.265
(0.005)
(0.004)
-0.001
-0.006
(0.005)
(0.003)
0.067
0.061
(0.007)
(0.006)
0.043
0.039
(0.001)
(0.001)
0.033
0.032
(0.000)
(0.000)
-0.059
-0.055
(0.001)
(0.001)
-0.042
-0.035
(0.002)
(0.002)
0.048
0.051
(0.005)
(0.010)
0.038
-0.010
(0.002)
(0.004)
No
Yes
Yes
Yes
2160012
2160012
0.331
0.396
College
(3)
(4)
-0.238
-0.257
(0.009)
(0.012)
0.041
-0.003
(0.012)
(0.009)
0.022
0.039
(0.010)
(0.012)
0.067
0.060
(0.002)
(0.002)
0.042
0.038
(0.001)
(0.001)
-0.062
-0.053
(0.002)
(0.002)
0.003
0.018
(0.005)
(0.004)
0.069
0.100
(0.010)
(0.018)
0.039
-0.023
(0.003)
(0.006)
No
Yes
Yes
Yes
553611
553611
0.420
0.498
on firm. p < 0.1, p < 0.05, p < 0.01.
28
29
(2)
Prof
-0.118
(0.021)
-0.030
(0.023)
-0.066
(0.023)
0.031
(0.002)
-0.047
(0.003)
0.042
(0.002)
0.024
(0.008)
0.046
(0.013)
0.035
(0.004)
Yes
75494
0.316
(3)
Tech
-0.202
(0.015)
0.027
(0.015)
-0.007
(0.017)
0.032
(0.001)
-0.051
(0.003)
0.036
(0.003)
-0.005
(0.007)
0.047
(0.015)
0.039
(0.006)
Yes
188774
0.352
(4)
Clerk
-0.212
(0.013)
-0.031
(0.012)
0.078
(0.014)
0.030
(0.001)
-0.047
(0.003)
0.026
(0.002)
-0.060
(0.008)
0.052
(0.008)
0.027
(0.003)
Yes
93251
0.209
(5)
Service
-0.336
(0.019)
0.037
(0.019)
0.042
(0.023)
0.036
(0.002)
-0.061
(0.004)
0.021
(0.003)
-0.045
(0.015)
0.003
(0.011)
0.029
(0.005)
Yes
36272
0.343
(6)
Craft
-0.278
(0.012)
-0.006
(0.011)
0.055
(0.014)
0.035
(0.001)
-0.065
(0.002)
0.036
(0.001)
-0.051
(0.004)
0.039
(0.008)
0.040
(0.004)
Yes
291776
0.215
Table 13: Gender Wage Gap Occupation Split

(7)
Mchn
-0.295
(0.012)
-0.016
(0.008)
0.098
(0.014)
0.027
(0.001)
-0.049
(0.001)
0.027
(0.001)
-0.048
(0.003)
0.035
(0.007)
0.049
(0.004)
Yes
511977
0.253
Industry*year
Observations
Adjusted R2
Size
Centrality
Children
Education
Experience2
Experience
Female*Exporter
Exporter
Female
(1)
Mng
-0.246
(0.017)
0.028
(0.013)
0.061
(0.018)
0.038
(0.002)
-0.060
(0.004)
0.048
(0.001)
0.063
(0.011)
0.091
(0.010)
0.051
(0.004)
Yes
133908
0.337
(8)
Elem
-0.237
(0.019)
-0.023
(0.015)
-0.003
(0.023)
0.027
(0.002)
-0.048
(0.004)
0.015
(0.002)
-0.071
(0.010)
0.043
(0.012)
0.044
(0.006)
Yes
51637
0.192
Table 14: Allowing for differential returns to all worker characteristics
Female
Exporter
Female*Exporter
Experience
Experience*Female
(1)
-0.133
(0.016)
0.010
(0.005)
0.031
(0.006)
0.035
(0.000)
-0.008
(0.001)
Experience*Exporter
Experience*Female*Exporter
Experience2
Experience2 *Female
-0.062
(0.001)
0.023
(0.001)
Experience2 *Exporter
Experience2 *Female*Exporter
Education
Education*Female
0.058
(0.001)
-0.005
(0.001)
Education*Exporter
Education*Female*Exporter
Children
Children*Female
-0.007
(0.002)
-0.092
(0.006)
Children*Exporter
Children*Female*Exporter
Centrality
Centrality*Female
0.050
(0.007)
0.018
(0.007)
Centrality*Exporter
Centrality*Female*Exporter
Size
Size*Female
0.036
(0.002)
0.013
(0.003)
Size*Exporter
Size*Female*Exporter
30
(2)
-0.277
(0.005)
-0.170
(0.037)
0.065
(0.007)
0.036
(0.001)
(3)
-0.171
(0.027)
-0.171
(0.035)
0.075
(0.033)
0.037
(0.001)
-0.008
(0.001)
-0.002
-0.002
(0.001)
(0.001)
-0.001
(0.001)
-0.063
-0.067
(0.001)
(0.001)
0.029
(0.003)
0.006
0.006
(0.002)
(0.002)
-0.007
(0.003)
0.044
0.044
(0.002)
(0.002)
-0.000
(0.002)
0.015
0.017
(0.002)
(0.002)
-0.006
(0.002)
-0.048
-0.019
(0.003)
(0.003)
-0.133
(0.008)
0.028
0.015
(0.005)
(0.004)
0.054
(0.010)
0.066
0.058
(0.006)
(0.006)
0.034
(0.009)
-0.013
-0.009
(0.009)
(0.009)
-0.020
(0.011)
0.045
0.044
(0.003)
(0.003)
0.005
(0.003)
-0.008
-0.010
(0.003)
(0.003)
0.011
(continued on next page)
Table 14: (continued)
Industry*year
Observations
Adjusted R2
Yes
2713623
0.416
Yes
2713623
0.416
(0.005)
Yes
2713623
0.417
on firm. p < 0.1, p < 0.05, p < 0.01.
31
Table 15: Paternity leave and the Gender Wage Gap
Exporter
Female*Exporter
Female*Exporter*PL
Exporter*PL
Profitability
Education
Experience
Experience2
Children
Centrality
Size
Spell FE
Female*year
Industry*year
Observations
Adjusted R2
All
(1)
-0.014
(0.005)
-0.029
(0.007)
0.182
(0.036)
0.133
(0.029)
0.006
(0.001)
-0.001
(0.001)
0.063
(0.002)
0.021
(0.008)
-0.054
(0.001)
-0.067
(0.006)
0.001
(0.007)
0.046
(0.006)
Yes
Yes
Yes
2713623
0.750
All Workers
Non
Coll
(2)
(3)
-0.010
-0.014
(0.005)
(0.008)
-0.026
-0.047
(0.007)
(0.014)
0.192
0.127
(0.039)
(0.067)
0.106
0.159
(0.027)
(0.042)
0.006
0.006
(0.001)
(0.001)
-0.001
0.001
(0.001)
(0.001)
0.075
0.063
(0.003)
(0.008)
0.042
0.014
(0.010)
(0.011)
-0.051
-0.065
(0.001)
(0.002)
-0.084
-0.035
(0.006)
(0.009)
-0.004
0.010
(0.008)
(0.007)
0.045
0.039
(0.006)
(0.008)
Yes
Yes
Yes
Yes
Yes
Yes
2160012
553611
0.696
0.813

All
Non
Coll
(4)
(5)
(6)
-0.009
-0.005
-0.013
(0.006)
(0.005)
(0.008)
-0.039 -0.033
-0.063
(0.010)
(0.010)
(0.018)
0.249
0.250
0.220
(0.052)
(0.058)
(0.090)
0.102
0.071
0.132
(0.032)
(0.032)
(0.049)
0.005
0.005
0.005
(0.001)
(0.001)
(0.001)
-0.000
-0.001
0.002
(0.001)
(0.001)
(0.001)
0.069
0.083
0.062
(0.003)
(0.003)
(0.008)
0.009
0.030
-0.009
(0.011)
(0.012)
(0.013)
-0.065
-0.074
-0.047
(0.004)
(0.005)
(0.005)
-0.076
-0.096
-0.043
(0.006)
(0.006)
(0.009)
-0.000
-0.005
0.008
(0.008)
(0.008)
(0.008)
0.045
0.046
0.031
(0.006)
(0.006)
(0.008)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
1621583
1267545
354038
0.733
0.688
0.787
PL is year specific and is defined as the share of men on paternity leave taking at least 8 weeks of leave.
Dummy for missing profitability included. Standard errors in parenthesis are clustered on firm.
p < 0.1, p < 0.05, p < 0.01.
32
Table 16: Kindergarten and the Gender Wage Gap
Exporter
Female*Exporter
Female*Exporter*KG
Female*KG
Exporter*KG
Profitability
KG
Education
Experience
Experience2
Children
Centrality
Size
Spell FE
Female*year
County*year
Industry*year
Observations
Adjusted R2
All
(1)
-0.053
(0.016)
-0.149
(0.028)
0.002
(0.000)
-0.002
(0.000)
0.001
(0.000)
0.005
(0.001)
-0.000
(0.001)
-0.001
(0.000)
0.073
(0.003)
0.034
(0.008)
-0.052
(0.001)
-0.068
(0.007)
0.010
(0.006)
0.048
(0.008)
Yes
Yes
Yes
Yes
1896790
0.757
All Workers
Non
(2)
-0.042
(0.015)
-0.151
(0.029)
0.002
(0.000)
-0.002
(0.000)
0.001
(0.000)
0.005
(0.001)
-0.001
(0.001)
-0.001
(0.000)
0.088
(0.004)
0.042
(0.009)
-0.049
(0.001)
-0.087
(0.007)
0.010
(0.008)
0.046
(0.007)
Yes
Yes
Yes
Yes
1485425
0.700
Coll
(3)
-0.051
(0.034)
-0.171
(0.070)
0.002
(0.001)
-0.002
(0.001)
0.001
(0.000)
0.006
(0.002)
0.001
(0.001)
-0.000
(0.000)
0.057
(0.009)
0.029
(0.015)
-0.060
(0.002)
-0.034
(0.012)
0.005
(0.010)
0.044
(0.011)
Yes
Yes
Yes
Yes
411365
0.817

All
Non
Coll
(4)
(5)
(6)
-0.029
-0.015
-0.036
(0.018)
(0.018)
(0.036)
-0.189 -0.196
-0.186
(0.040)
(0.042)
(0.085)
0.002
0.003
0.002
(0.001)
(0.001)
(0.001)
-0.002
-0.002
-0.002
(0.001)
(0.001)
(0.001)
0.000
0.000
0.001
(0.000)
(0.000)
(0.000)
0.005
0.004
0.006
(0.001)
(0.001)
(0.002)
0.000
-0.000
0.002
(0.001)
(0.001)
(0.002)
-0.000
-0.000
-0.000
(0.000)
(0.000)
(0.000)
0.080
0.097
0.057
(0.003)
(0.004)
(0.009)
0.020
0.028
0.006
(0.011)
(0.011)
(0.019)
-0.054
-0.065
-0.033
(0.004)
(0.005)
(0.006)
-0.076
-0.096
-0.040
(0.007)
(0.007)
(0.012)
0.006
0.007
-0.001
(0.008)
(0.010)
(0.012)
0.046
0.044
0.039
(0.008)
(0.007)
(0.010)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
1099014
842514
256500
0.736
0.689
0.787
Notes: Estimates are based on the panel of worker-level data for 2000-2010. Dependent variable: log
wage. KG is year and municipality specific and is defined as the share of children between the ages of
1 and 5 enrolled in kindergarten. Dummy for missing profitability included. Norway has 19 counties and
439 municipalities. Standard errors in parenthesis are clustered on firm. p < 0.1, p < 0.05, p < 0.01.
33
ERIA Discussion Paper Series
No.
Author(s)
2015-25
Esther Ann BLER,

Beata JAVORCIK,
Karen Helene
ULLTVEI-MOE
Tristan Leo Dallo

2015-24 AGUSTIN and Martin
SCHRDER
Title
Globalization: A Womans Best Friend?

Exporters and the Gender Wage Gap
Year
Mar
2015
Mar
The Indian Automotive Industry and the ASEAN
Supply Chain Relations
2015
Mar
Hideo KOBAYASHI
and Yingshan JIN
The CLMV Automobile and Auto Parts Industry
2015-22
Hideo KOBAYASHI
Mar
Current State and Issues of the Automobile and
Auto Parts Industries in ASEAN
2015
2015-21
Yoshifumi
FUKUNAGA
Mar
Assessing the Progress of ASEAN MRAs on
Professional Services
2015
2015-20
Yoshifumi
FUKUNAGA and
Hikari ISHIDO
Mar
Values and Limitations of the ASEAN
Agreement on the Movement of Natural Persons 2015
Nanda NURRIDZKI
Mar
Learning from the ASEAN + 1 Model and the
ACIA
2015
2015-23
2015-19
Patarapong
INTARAKUMNERD
2015-18 and Pun-Arj
CHAIRATANA and
Preeda CHAYANAJIT
2015-17
Rajah RASIAH and

Yap Xiao SHAN
2015-16
Rajah RASIAH and

Yap Xiao SHAN
2015-15
Xin Xin KONG, Miao

ZHANG and Santha
Chenayah RAMU
2015
Global Production Networks and Host-Site Feb

Industrial Upgrading: The Case of the
2015
Semiconductor Industry in Thailand
Institutional Support, Regional Trade Linkages Feb
and Technological Capabilities in the
2015
Semiconductor Industry in Singapore
Institutional Support, Regional Trade Linkages Feb
and Technological Capabilities in the
2015
Semiconductor Industry in Malaysia
Feb
Chinas Semiconductor Industry in Global Value
Chains
2015
34
No.
Author(s)
2015-14
Tin Htoo NAING and

Yap Su FEI
2015-13
Vanthana NOLINTHA
and Idris JAJRI
Miao ZHANG, Xin Xin

KONG, Santha
Chenayah RAMU
NGUYEN Dinh Chuc,
NGUYEN Dinh Anh,
2015-11 NGUYEN Ha Trang
and NGUYEN Ngoc
Minh
Pararapong
2015-10 INTERAKUMNERD
and Kriengkrai
TECHAKANONT
2015-12
2015-09
Rene E. OFRENEO
Rajah RASIAH, Rafat

2015-08 Beigpoor
SHAHRIVAR, Abdusy
Syakur AMIN
Yansheng LI, Xin Xin
2015-07 KONG, and Miao
ZHANG
2015-06
Mukul G. ASHER and

Fauziah ZEN
Lili Yan ING, Stephen

MAGIERA, and Anika
WIDIANA
Gemma ESTRADA,
James ANGRESANO,
Jo Thori LIND, Niku
MTNEN, William
2015-04 MCBRIDE, Donghyun
PARK, Motohiro
SATO, and Karin
SVANBORGSJVALL
2015-05
Title
Year
Feb
Multinationals, Technology and Regional
Linkages in Myanmars Clothing Industry
2015
The Garment Industry in Laos: Technological Feb
Capabilities, Global Production Chains and
2015
Competitiveness
Feb
The Transformation of the Clothing Industry in
China
2015
Host-site institutions, Regional Production Feb
Linkages and Technological Upgrading: A study
2015
of Automotive Firms in Vietnam
Intra-industry Trade, Product Fragmentation and
Technological Capability Development in Thai
Automotive Industry
Auto and Car Parts Production: Can the
Philippines Catch Up with Asia
Feb
2015
Feb
2015
Host-site Support, Foreign Ownership, Regional

Linkages and Technological Capabilites:
Evidence from Automotive Firms in Indonesia
Feb
Industrial Upgrading in Global Production

Networks: Te Case of the Chinese Automotive
Industry
Feb
Social Protection in ASEAN: Challenges and

Initiatives for Post-2015 Vision
Business Licensing: A Key to Investment
Climate Reform
Fiscal Policy and Equity in Advanced

Economies: Lessons for Asia
35
2015
2015
Feb
2015
Feb
2015
Jan
2015
No.
2015-03
Author(s)
Erlinda M. MEDALLA
Title
Towards an Enabling Set of Rules of Origin for

the Regional Comprehensive Economic
Partnership
Year
Jan
2015
Archanun
2015-02 KOHPAIBOON and
Juthathip
JONGWANICH
Use of FTAs from Thai Experience
2015-01
Misa OKABE
Impact of Free Trade Agreements on Trade in

East Asia
Hikari ISHIDO
Coverage of Trade in Services under ASEAN+1

FTAs
Junianto James
LOSARI
Searching for an Ideal International Investment

Protection Regime for ASEAN + Dialogue
Partners (RCEP): Where Do We Begin?
Dec
Dayong ZHANG and

David C. Broadstock
Impact of International Oil Price Shocks on

Consumption Expenditures in ASEAN and East
Asia
Nov
Dandan ZHANG,
Xunpeng SHI, and Yu
SHENG
Enhanced Measurement of Energy Market

Integration in East Asia: An Application of
Dynamic Principal Component Analysis
Nov
Yanrui WU
Deregulation, Competition, and Market

Integration in Chinas Electricity Sector
2014-26
2014-25
2014-24
2014-23
2014-22
2014-21 Yanfei LI and Youngho

CHANG
Yu SHENG, Yanrui
WU, Xunpeng SHI,
Dandan ZHANG
Andindya
2014-19 BHATTACHARYA
and Tania
BHATTACHARYA
2014-20
Jan
Infrastructure Investments for Power Trade and

Transmission in ASEAN+2: Costs, Benefits,
Long-Term
Contracts,
and
Prioritised
Development
Market Integration and Energy Trade Efficiency:
An Application of Malmqviat Index to Analyse
Multi-Product Trade
2015
Jan
2015
Dec
2014
2014
2014
2014
Nov
2014
Nov
2014
Nov
2014
Nov
ASEAN-India Gas Cooperation: Redifining

Indias Look East Policy with Myanmar
2014
Sep
Olivier CADOT, Lili

Yan ING
How Restrictive Are ASEANs RoO?
2014-17
Sadayuki TAKII
Import Penetration, Export Orientation, and Plant

Size in Indonesian Manufacturing
2014
2014-16
Tomoko INUI, Keiko

ITO, and Daisuke
MIYAKAWA
Japanese Small and Medium-Sized Enterprises July

Export Decisions: The Role of Overseas Market
2014
Information
2014-18
2014
July
36
No.
Author(s)
Title
Year
Han PHOUMIN and

Fukunari KIMURA
Trade-off
Relationship
between
Energy
Intensity-thus energy demand- and Income June
Level: Empirical Evidence and Policy
Implications for ASEAN and East Asia 2014
Countries
Cassey LEE
The Exporting and Productivity Nexus: Does

Firm Size Matter?
2014
2014-13
Yifan ZHANG
Productivity Evolution of Chinese large and

Small Firms in the Era of Globalisation
2014
2014-12
Valria SMEETS,
Sharon
TRAIBERMAN,
Frederic WARZYNSKI
Offshoring and the Shortening of the Quality

Ladder:Evidence from Danish Apparel
2014
May
Inkyo CHEONG
Koreas Policy Package for Enhancing its FTA

Utilization and Implications for Koreas Policy
Sothea OUM, Dionisius Constraints, Determinants of SME Innovation,

NARJOKO, and
and the Role of Government Support
Charles HARVIE
Migrant Networks and Trade: The Vietnamese
2014-09 Christopher PARSONS
and Pierre-Louis Vzina Boat People as a Natural Experiment
May
2014-15
May
2014-14
May
2014-11
May
2014
2014-10
Kazunobu
HAYAKAWA and
Toshiyuki
MATSUURA
2014
May
2014
Dynamic Tow-way Relationship between

Exporting and Importing: Evidence from Japan
May
2014
2014-07 DOAN Thi Thanh Ha

and Kozo KIYOTA
Firm-level Evidence on Productivity

Differentials and Turnover in Vietnamese
Manufacturing
Apr
2014-06 Larry QIU and Miaojie

YU
Multiproduct Firms, Export Product Scope, and

Trade Liberalization: The Role of Managerial
Efficiency
Apr
Han PHOUMIN and

Shigeru KIMURA
Analysis on Price Elasticity of Energy Demand

in East Asia: Empirical Evidence and Policy
Implications for ASEAN and East Asia
Apr
Youngho CHANG and

Yanfei LI
Non-renewable Resources in Asian Economies:

Perspectives of Availability, Applicability,
Acceptability, and Affordability
Feb
Yasuyuki SAWADA
and Fauziah ZEN
Disaster Management in ASEAN
2014-08
2014-05
2014-04
2014-03
2014
2014
2014
2014
Jan
2014
37
No.
Author(s)
Title
Year
Jan
2014-02
2014-01
2013-38
2013-37
2013-36
2013-35
2013-34
2013-33
2013-32
2013-31
2013-30
2013-29
2013-28
2013-27
Cassey LEE
Competition Law Enforcement in Malaysia

Jan
Rizal SUKMA
ASEAN Beyond 2015: The Imperatives for

Further Institutional Changes
Toshihiro OKUBO,
Fukunari KIMURA,
Nozomu TESHIMA
Asian Fragmentation in the Global Financial

Crisis
Dec
Xunpeng SHI and

Cecilya MALIK
Assessment of ASEAN Energy Cooperation

within the ASEAN Economic Community
Dec
Tereso S. TULLAO, Jr.

And Christopher James
CABUAY
Eduction and Human Capital Development to

Strengthen R&D Capacity in the ASEAN
Dec
Dec
Paul A. RASCHKY
Estimating the Effects of West Sumatra Public

Asset Insurance Program on Short-Term
Recovery after the September 2009 Earthquake
2014
Nipon
Impact of the 2011 Floods, and Food
POAPONSAKORN and Management in Thailand
Pitsom MEETHOM
2014
2013
2013
2013
2013
Nov
2013
Nov
Mitsuyo ANDO
Development and Resructuring of Regional

Production/Distribution Networks in East Asia
Mitsuyo ANDO and

Fukunari KIMURA
Evolution of Machinery Production Networks:

Linkage of North America with East Asia?
Nov
Mitsuyo ANDO and

Fukunari KIMURA
What are the Opportunities and Challenges for

ASEAN?
Nov
Nov
Simon PEETMAN
Standards Harmonisation in ASEAN: Progress,

Challenges and Moving Beyond 2015
Jonathan KOH and

Andrea Feldman
MOWERMAN
Towards a Truly Seamless Single Windows and

Trade Facilitation Regime in ASEAN Beyond
2015
Nov
Nov
Rajah RASIAH
Stimulating Innovation in ASEAN Institutional

Support, R&D Activity and Intelletual Property
Rights
Maria Monica
WIHARDJA
Financial Integration Challenges in ASEAN

beyond 2015
Nov
38
2013
2013
2013
2013
2013
2013
2013
No.
Author(s)
Tomohiro
2013-26 MACHIKITA and
Yasushi UEKI
2013-25
Fukunari KIMURA
Olivier CADOT
Ernawati MUNADI
2013-24
Lili Yan ING
Title
Who Disseminates Technology to Whom, How,

and Why: Evidence from Buyer-Seller Business
Networks
Nov
Reconstructing the Concept of Single Market a

Production Base for ASEAN beyond 2015
Oct
Streamlining NTMs in ASEAN:
Oct
The Way Forward
2013
Charles HARVIE,
2013-23
Year
Small and Medium Enterprises Access to

Dionisius NARJOKO, S Finance: Evidence from Selected Asian
Economies
othea OUM
2013
2013
Oct
2013
Toward a Single Aviation Market in ASEAN:

Regulatory Reform and Industry Challenges
Oct
Shintaro SUGIYAMA,
2013-21
Fauziah ZEN
Moving MPAC Forward: Strengthening PublicPrivate Partnership, Improving Project Portfolio

and in Search of Practical Financing Schemes
Oct
Barry DESKER, Mely

CABALLERO2013-20
ANTHONY, Paul
TENG
Thought/Issues Paper on ASEAN Food Security:

Towards a more Comprehensive Framework
Oct
Toshihiro KUDO,
2013-19 Satoru KUMAGAI, So
UMEZAKI
Making Myanmar the Star Growth Performer in

ASEAN in the Next Decade: A Proposal of Five
Growth Strategies
Sep
Managing Economic Shocks and

Macroeconomic Coordination in an Integrated
Region: ASEAN Beyond 2015
Sep
2013-22
Alan Khee-Jin TAN
2013
Hisanobu SHISHIDO,
2013-18
2013-17
2013-16
2013-15
Ruperto MAJUCA
Cassy LEE and Yoshifumi Competition Policy Challenges of Single Market

FUKUNAGA
and Production Base
Simon TAY
Danilo C. ISRAEL and
2013
2013
2013
2013
Sep
2013
Growing an ASEAN Voice? : A Common

Platform in Global and Regional Governance
Sep
Impacts of Natural Disasters on Agriculture, Food

Security, and Natural Resources and Environment in
Aug
39
2013
No.
Author(s)
Title
Year
Roehlano M. BRIONES
the Philippines
2013-14
Allen Yu-Hung LAI and

Seck L. TAN
Impact of Disasters and Disaster Risk Management in Aug

Singapore: A Case Study of Singapores Experience
2013
in Fighting the SARS Epidemic
Brent LAYTON
Impact of Natural Disasters on Production Networks

and Urbanization in New Zealand
Aug
2013-13
Impact of Recent Crises and Disasters on Regional

Production/Distribution Networks and Trade in Japan
Aug
Economic and Welfare Impacts of Disasters in East

Asia and Policy Responses: The Case of Vietnam
Aug
Sann VATHANA, Sothea

OUM, Ponhrith KAN,
Colas CHERVIER
Impact of Disasters and Role of Social Protection in

Natural Disaster Risk Management in Cambodia
Aug
Sommarat
CHANTARAT, Krirk
PANNANGPETCH,
Nattapong
PUTTANAPONG,
Preesan RAKWATIN,
and Thanasin
TANOMPONGPHANDH
Index-Based Risk Financing and Development of

Natural Disaster Insurance Programs in Developing
Asian Countries
Aug
Ikumo ISONO and Satoru

KUMAGAI
Long-run Economic Impacts of Thai Flooding:

Geographical Simulation Analysis
July
Yoshifumi FUKUNAGA
and Hikaru ISHIDO
Assessing the Progress of Services Liberalization in

the ASEAN-China Free Trade Area (ACFTA)
May
Ken ITAKURA,
Yoshifumi FUKUNAGA,
and Ikumo ISONO
A CGE Study of Economic Impact of Accession of

Hong Kong to ASEAN-China Free Trade Agreement
May
Misa OKABE and Shujiro

URATA
The Impact of AFTA on Intra-AFTA Trade
Kohei SHIINO
How Far Will Hong Kongs Accession to ACFTA

will Impact on Trade in Goods?
2013-12
2013-11
2013-10
2013-09
2013-08
2013-07
2013-06
2013-05
2013-04
Mitsuyo ANDO
Le Dang TRUNG
2013
2013
2013
2013
2013
2013
2013
2013
2013
May
2013
40
May
2013
No.
Author(s)
Title
2013-03
Cassey LEE and

Yoshifumi FUKUNAGA
ASEAN Regional Cooperation on Competition

Policy
Yoshifumi FUKUNAGA
and Ikumo ISONO
Taking ASEAN+1 FTAs towards the RCEP:
Year
Apr
2013
Jan
2013-02
2013
A Mapping Study
2013-01
Ken ITAKURA
Impact of Liberalization and Improved Connectivity

Jan
and Facilitation in ASEAN for the ASEAN Economic
2013
Community
2012-17
Sun XUEGONG, Guo

LIYAN, Zeng ZHENG
Market Entry Barriers for FDI and Private Investors:

Lessons from Chinas Electricity Market
Aug
2012
2012-16
Yanrui WU
Electricity Market Integration: Global Trends and

Implications for the EAS Region
Aug
2012
2012-15
Youngho CHANG,
Yanfei LI
Power Generation and Cross-border Grid Planning

for the Integrated ASEAN Electricity Market: A
Dynamic Linear Programming Model
Aug
2012
2012-14
Yanrui WU, Xunpeng

SHI
Economic Development, Energy Market Integration

and Energy Demand: Implications for East Asia
Aug
2012
2012-13
Joshua AIZENMAN,
Minsoo LEE, and
Donghyun PARK
The Relationship between Structural Change and

Inequality: A Conceptual Overview with Special
Reference to Developing Asia
July
2012
2012-12
Hyun-Hoon LEE, Minsoo

LEE, and Donghyun
PARK
Growth Policy and Inequality in Developing Asia:

Lessons from Korea
July
2012
2012-11
Cassey LEE
Knowledge Flows, Organization and Innovation:

Firm-Level Evidence from Malaysia
June
2012
2012-10
Jacques MAIRESSE,
Pierre MOHNEN, Yayun
ZHAO, and Feng ZHEN
Globalization, Innovation and Productivity in

Manufacturing Firms: A Study of Four Sectors of
China
June
2012
2012-09
Ari KUNCORO
Globalization and Innovation in Indonesia: Evidence

from Micro-Data on Medium and Large
Manufacturing Establishments
June
2012
2012-08
Alfons
PALANGKARAYA
The Link between Innovation and Export: Evidence

from Australias Small and Medium Enterprises
June
2012
2012-07
Chin Hee HAHN and

Chang-Gyun PARK
Direction of Causality in Innovation-Exporting

Linkage: Evidence on Korean Manufacturing
June
2012
41
No.
Author(s)
Title
Year
2012-06
Keiko ITO
Source of Learning-by-Exporting Effects: Does

Exporting Promote Innovation?
June
2012
2012-05
Rafaelita M. ALDABA
Trade Reforms, Competition, and Innovation in the

Philippines
June
2012
The Role of Trade Costs in FDI Strategy of

Heterogeneous Firms: Evidence from Japanese Firmlevel Data
June
2012-04
Toshiyuki MATSUURA
and Kazunobu
HAYAKAWA
2012-03
Kazunobu HAYAKAWA,
How Does Country Risk Matter for Foreign Direct
Fukunari KIMURA, and
Investment?
Hyun-Hoon LEE
2012-02
Ikumo ISONO, Satoru

KUMAGAI, Fukunari
KIMURA
2012-01
Mitsuyo ANDO and

Fukunari KIMURA
2011-10
Tomohiro MACHIKITA
and Yasushi UEKI
2011-09
Joseph D. ALBA, WaiMun CHIA, and

Donghyun PARK
Agglomeration and Dispersion in China and ASEAN:

A Geographical Simulation Analysis
How Did the Japanese Exports Respond to Two Crises
in the International Production Network?: The Global
Financial Crisis and the East Japan Earthquake
Interactive Learning-driven Innovation in UpstreamDownstream Relations: Evidence from Mutual
Exchanges of Engineers in Developing Economies
Foreign Output Shocks and Monetary Policy
Regimes in Small Open Economies: A DSGE
Evaluation of East Asia
2012
Feb
2012
Jan
2012
Jan
2012
Dec
2011
Dec
2011
Tomohiro MACHIKITA
and Yasushi UEKI
Impacts of Incoming Knowledge on Product

Innovation: Econometric Case Studies of Technology
Transfer of Auto-related Industries in Developing
Economies
2011-07
Yanrui WU
Gas Market Integration: Global Trends and

Implications for the EAS Region
Nov
2011
Philip Andrews-SPEED
Energy Market Integration in East Asia: A Regional

Public Goods Approach
Nov
2011-06
Energy Market Integration and Economic

Convergence: Implications for East Asia
Oct
2011
2011-08
Yu SHENG,
2011-05
Xunpeng SHI
2011-04
Sang-Hyop LEE, Andrew

MASON, and Donghyun
PARK
Why Does Population Aging Matter So Much for

Asia? Population Aging, Economic Security and
Xunpeng SHI,
Harmonizing Biodiesel Fuel Standards in East Asia:

Current Status, Challenges and the Way Forward
2011-03
Shinichi GOTO
Nov
2011
2011
Aug
2011
Economic Growth in Asia
42
May
2011
No.
2011-02
2011-01
Author(s)
Title
Year
Liberalization of Trade in Services under ASEAN+n :

A Mapping Exercise
May
2011
Location Choice of Multinational Enterprises in

China: Comparison between Japan and Taiwan
Mar
2011
Hikari ISHIDO
Kuo-I CHANG,
Kazunobu HAYAKAWA
Toshiyuki MATSUURA
2010-11
Charles HARVIE,
Dionisius NARJOKO,
Sothea OUM
Firm Characteristic Determinants of SME

Participation in Production Networks
Oct
2010
2010-10
Mitsuyo ANDO
Machinery Trade in East Asia, and the Global

Financial Crisis
Oct
2010
International Production Networks in Machinery

Industries: Structure and Its Evolution
Sep
2010
2010-08
Tomohiro MACHIKITA,
Shoichi MIYAHARA,
Masatsugu TSUJI, and
Yasushi UEKI
Detecting Effective Knowledge Sources in Product

Innovation: Evidence from Local Firms and
MNCs/JVs in Southeast Asia
Aug
2010
2010-07
Tomohiro MACHIKITA,
Masatsugu TSUJI, and
Yasushi UEKI
How ICTs Raise Manufacturing Performance: Firmlevel Evidence in Southeast Asia
Aug
2010
2010-06
Xunpeng SHI
Carbon Footprint Labeling Activities in the East Asia

Summit Region: Spillover Effects to Less Developed
Countries
July
2010
2010-05
Kazunobu HAYAKAWA,
Fukunari KIMURA
2010-09
Ayako OBASHI
Firm-level Analysis of Globalization: A Survey of

the Eight Literatures
Mar
2010
Tomohiro MACHIKITA
Tomohiro MACHIKITA
2010-04
and Yasushi UEKI
The Impacts of Face-to-face and Frequent

Interactions on Innovation:
Upstream-Downstream Relations
Tomohiro MACHIKITA
Innovation in Linked and Non-linked Firms:
and Yasushi UEKI
Effects of Variety of Linkages in East Asia
2010-03
Tomohiro MACHIKITA
2010-02
and Yasushi UEKI
Feb
2010
Search-theoretic Approach to Securing New

Suppliers: Impacts of Geographic Proximity for
Importer and Non-importer
43
Feb
2010
Feb
2010
No.
Author(s)
Tomohiro MACHIKITA
2010-01
and Yasushi UEKI
2009-23
Dionisius NARJOKO
Title
Spatial Architecture of the Production Networks in
Southeast Asia:
Year
Feb
2010
Empirical Evidence from Firm-level Data

Foreign Presence Spillovers and Firms Export
Response:
Nov
2009
Evidence from the Indonesian Manufacturing
2009-22
Kazunobu HAYAKAWA,
Daisuke HIRATSUKA,
Who Uses Free Trade Agreements?
Kohei SHIINO, and Seiya
SUKEGAWA
2009-21
Ayako OBASHI
Resiliency of Production Networks in Asia:

Evidence from the Asian Crisis
Oct
2009
2009-20
Mitsuyo ANDO and

Fukunari KIMURA
Fragmentation in East Asia: Further Evidence
Oct
2009
2009-19
Xunpeng SHI
The Prospects for Coal: Global Experience and

Implications for Energy Policy
Sept
2009
2009-18
Sothea OUM
Income Distribution and Poverty in a CGE

Framework: A Proposed Methodology
Jun
2009
2009-17
Erlinda M. MEDALLA
and Jenny BALBOA
ASEAN Rules of Origin: Lessons and

Recommendations for the Best Practice
Jun
2009
2009-16
Masami ISHIDA
Special Economic Zones and Economic Corridors
Jun
2009
2009-15
Toshihiro KUDO
Border Area Development in the GMS: Turning the

Periphery into the Center of Growth
May
2009
2009-14
Claire HOLLWEG and

Marn-Heong WONG
Measuring Regulatory Restrictions in Logistics

Services
Apr
2009
2009-13
Loreli C. De DIOS
Business View on Trade Facilitation
Apr
2009
2009-12
Patricia SOURDIN and

Richard POMFRET
Monitoring Trade Costs in Southeast Asia
Apr
2009
Barriers to Trade in Health and Financial Services in

ASEAN
Apr
2009
Philippa DEE and

2009-11
Huong DINH
2009-10
Sayuri SHIRAI
Nov
2009
The Impact of the US Subprime Mortgage Crisis on

Apr
the World and East Asia: Through Analyses of Cross2009
border Capital Movements
44
No.
Author(s)
Title
Year
Akie IRIYAMA
International Production Networks and Export/Import

Responsiveness to Exchange Rates: The Case of
Japanese Manufacturing Firms
Mar
2009
2009-08
Archanun
KOHPAIBOON
Vertical and Horizontal FDI Technology

Spillovers:Evidence from Thai Manufacturing
Mar
2009
2009-07
Kazunobu HAYAKAWA,
Gains from Fragmentation at the Firm Level:
Evidence from Japanese Multinationals in East Asia
Toshiyuki MATSUURA
Mar
2009
2009-06
Dionisius A. NARJOKO
Plant Entry in a More

LiberalisedIndustrialisationProcess: An Experience
of Indonesian Manufacturing during the 1990s
Mar
2009
2009-05
Kazunobu HAYAKAWA,
Firm-level Analysis of Globalization: A Survey
Tomohiro MACHIKITA
Mar
2009
2009-04
Chin Hee HAHN and

Chang-Gyun PARK
Learning-by-exporting in Korean Manufacturing: A

Plant-level Analysis
Mar
2009
2009-03
Ayako OBASHI
Stability of Production Networks in East Asia:

Duration and Survival of Trade
Mar
2009
2009-02
Fukunari KIMURA
The Spatial Structure of Production/Distribution

Networks and Its Implication for Technology
Transfers and Spillovers
Mar
2009
2009-01
Fukunari KIMURA and

Ayako OBASHI
International Production Networks: Comparison

between China and ASEAN
Jan
2009
2008-03
Kazunobu HAYAKAWA
and Fukunari KIMURA
The Effect of Exchange Rate Volatility on

International Trade in East Asia
Dec
2008
2008-02
Satoru KUMAGAI,
Toshitaka GOKAN,
Ikumo ISONO, and
Souknilanh KEOLA
Predicting Long-Term Effects of Infrastructure

Development Projects in Continental South East
Asia: IDE Geographical Simulation Model
Dec
2008
2008-01
Kazunobu HAYAKAWA,
Firm-level Analysis of Globalization: A Survey
Tomohiro MACHIKITA
Dec
2008
Mitsuyo ANDO and

2009-09
45

Globalization: A Woman's Best Friend? Exporters and The Gender Wage Gap

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Globalization: A Woman's Best Friend? Exporters and The Gender Wage Gap

Caricato da

Copyright:

Formati disponibili

ERIA-DP-2015-25

ERIA Discussion Paper Series

Globalization: A Womans Best Friend? Exporters and

Karen Helene ULLTVEI-MOE

Keywords: Exporters, Globalization, Gender Wage Gap

Globalization: A Womans Best Friend?

Karen Helene Ulltveit-Moe

University of Oslo and ESOP.

University of Oxford, ESOP and CEPR.

University of Oslo, ESOP and CEPR.

Data and Basic Facts

Construction of the Sample and Variables

Some Basic Facts on Women and Exporters

Notes: Estimates are based on the panel of worker-level

p < 0.1, p < 0.05, p < 0.01.

Empirical Model and Baseline Results

Exporters and the Gender Wage Gap: Controlling for Observables

Table 2: Gender Wage Gap

Our reference group is males in non-exporting firms.

Exporters and the Gender Wage Gap: Controlling for Unobservables

No workers in our sample switch gender.

Table 3: Worker and Spell Fixed Effects

Women, Commitment and the Gender Wage Gap: Exploiting

Paternity Leave Reforms

Workers < 45 years old

There are 439 municipalities and 19 counties in Norway.

Table 5: Kindergarten and the Gender Wage Gap

Workers < 45 years old

Exporter Heterogeneity and the Gender Wage Gap

Addressing an Alternative Explanation

p < 0.1, p < 0.05, p < 0.01.

Irarrazabal, A., A. Moxnes, and K. H. Ulltveit-Moe (2013). Heterogeneous firms or heterogeneous

Do exporters really pay higher wages?

Figure 1: Share of exporters across industries

Table 10: Descriptive Statistics

Table 11: Exporter premia

0.032 0.036 0.050 0.032 0.049 0.035 0.048 0.053 0.050

Table 12: Gender Wage Gap Education Split

Table 13: Gender Wage Gap Occupation Split

Table 14: Allowing for differential returns to all worker characteristics

Table 14: (continued)

Table 15: Paternity leave and the Gender Wage Gap

Workers < 45 years old

p < 0.1, p < 0.05, p < 0.01.

Table 16: Kindergarten and the Gender Wage Gap

Workers < 45 years old

ERIA Discussion Paper Series

Esther Ann BLER,

Tristan Leo Dallo

Globalization: A Womans Best Friend?

The CLMV Automobile and Auto Parts Industry

Rajah RASIAH and

Rajah RASIAH and

Xin Xin KONG, Miao

Global Production Networks and Host-Site Feb

Tin Htoo NAING and

Miao ZHANG, Xin Xin

Rajah RASIAH, Rafat

Mukul G. ASHER and

Lili Yan ING, Stephen

Host-site Support, Foreign Ownership, Regional

Industrial Upgrading in Global Production

Social Protection in ASEAN: Challenges and